Superconducting digital technology has made rapid strides in recent years. Energy-efficient single flux quantum (SFQ) circuits have been proposed so far and complicated circuits like a microprocessor have been demonstrated at several tens of GHz. For example, our group successfully demonstrated 100 GHz operation of a bit-serial microprocessor based on the multiple-voltage SFQ (MV-SFQ) circuit. In addition, we designed a RISC-based MV-SFQ microprocessor with an embedded instruction and data memory, and confirmed execution of all the instructions stored in the instruction memory. Note that these circuits are designed based on the SFQ-specific CAD tools in a digital domain and sometimes based on a microwave simulator for analog components. Furthermore, the introduction of ferromagnetic materials to devices or circuits, the invention of superconducting nano-devices, and the combination with CMOS integrated circuits enhance flexibility of the analog circuits as well as digital circuits and lead to technological innovation. Almost all the remaining issues such as large capacity memories, rectifiers, will be overcome by using these technologies. Superconducting digital technology is now transformed into cryogenic digital technology. Although the above-mentioned technologies are based on Nb or NbN films, I will give some comments on the importance of high-temperature-superconductor (HTS) films. In fact, the waveguides made of HTS films are essential for, broadband communications with ultra-low heat inflow between a 1st stage and a 2nd stage of a cryocooler. HTS analog active nanodevices would be attractive for an amplifier operating around 50 K.

Superconductor integrated circuits (ICs), built with niobium Josephson junctions (JJs), offer an attractive combination of features for mixed-signal electronics up to frequencies above 100 GHz. In these circuits, conversion between analog and digital domains can be done with high fidelity, by exploiting magnetic flux quantization, and at high speed. Niobium JJs have picosecond switching times and enable sampling rates above 100 GHz. This fast and accurate analog-to-digital conversion function is strengthened by the ability to integrate ultrafast, low-power digital circuits and low-loss analog circuits on the same chip. These features give rise to a class of mixed-signal circuits for a diversity of applications, ranging from communications to quantum computing. Among these are software radio and digital radar applications which benefit from digital signal processing applied directly on digitized radio frequency (RF) waveforms. On the other hand, the same technology applies to reading out cryogenic detectors and qubits. In this paper, we describe the current state-of-the-art of digital-RF systems, featuring superconductor mixed-signal circuits. The latest generation of these robust cryocooled systems are modular and have been operated for a variety of applications over the last four years. In these systems, the function of superconductor mixed-signal ICs is augmented by semiconductor cryogenic and room-temperature electronics. We also describe the next generation of superconductor ICs that are faster and feature much higher integration density and complexity. Finally, microwave design challenges will be discussed in two critical areas: (1) interchip interconnects for analog signals and picosecond-wide single flux quantum (SFQ) pulses supporting rates above 100 GHz , and (2) generation, distribution and synchronization of clock signals also above 100 GHz.

In recent years, great efforts have been devoted to the research and development of high temperature superconducting (HTS) filters in China. Novel HTS filter design technologies were well developed including those in coupling matrix extraction, wide stop band realization, liner phase, multiplexer, multi-band and multi-mode designs, and frequency tunable filters. Based on these techniques, high performance HTS filters were constructed and excellent specifications were achieved, e.g., 0.05 dB minimum insertion loss, -23 dB return loss, -110 dB out-of-band rejection and 220 dB/MHz band-edge slope, etc., which are among the best results in the literatures. Applications of HTS filters are also successful. A demonstration wireless communication cluster with HTS filter subsystems was built in the urban area of Beijing. Each base transceiver station (BTS) was installed with an HTS filter subsystem consisting of six HTS filters with the same performance. The measurement results show a 2.35 dB decrease of mobile phone mean transmit power when the normal filter subsystems were replaced by HTS filter subsystems. Based on the success in constructing an HTS filter handling high power up to 11.7 W, one of the highest records in the literature, the world first HTS transceiver was constructed, which was installed in a 3rd generation TD-SCDMA base station (it works in TDD mode, i.e., the receiving and emitting signals use the same frequency channel). Field trial in commercial network showed excellent improvements in reducing bit error (80%), enhancing anti-interference (10 dBm), restricting spurious power in emitting signals (26 dBm), and also in receiving quality of video communications. Field trial for the 4th generation mobile communication systems (e.g. TD-LTE) is now being carried out and results will be reported in the symposium. Another interesting application is to radar system, e.g., the meteorological radar, which is sometimes paralyzed by heavy electromagnetic interference in urban area due to the lack of extremely narrow bandwidth pre-selective filters. Laboratory tests proved that, with HTS subsystem, improvements of sensitivity (3.8 dB) and interference rejection (48.4 dB) have been achieved. A demonstration HTS meteorological radar station was then set up in Beijing. Comparison measurements showed that due to interference, the conventional wind profiler could not attain stable and reliable wind profiles above 1500m where the echoes are weak. In contrast, the HTS wind profiler provided complete and coincident wind profiles up to 3500 m. In poor weather conditions or severe electromagnetic environment, the conventional wind profiler almost stopped working, while the HTS wind profiler functioned well, and complete wind profile data on both the time scale and the spatial scale could be obtained by using the HTS wind profiler. The third application is to radio astronomy and deep space detection. HTS filters were deployed in the ground stations of Chinese lunar exploration project, which played an important role in monitoring the mission of Chang'e No. 3 satellite. Finally, applications in space technology are most challenging and attractive. In 2005, the first ground test system showed that a reduction of 73% in noise temperature was obtained by substituting HTS front-end for its conventional counterpart. On October 14, 2012, the HTS experimental system was successfully launched into orbit as a payload of a civil experimental satellite for new technologies (Practice No 9). The received on-orbit experimental data showed that the HTS system worked perfect in the past years. This is the world second successful space experiment for HTS devices after the American’s HTSSE. Application of HTS filter front-end in a space science project is also ready for space mission with Chinese Space Laboratory, to be launched in mid September, 2016.

The Micro-Electro-Mechanical System (MEMS) technology has the potential of replacing many radio frequency (RF) and microwave components used in today's communication systems. In particular, the use of RF MEMS in the reconfigurable filters would not only reduce substantially the size, weight but also promise linearity performance that is far superior to other technologies. The ability to integrate MEMS with superconductor filters promise to realize tunable filters with an unprecedented performance. The talk will address the behaviour of RF MEMS switches at cryogenic temperatures illustrating their use in the realization of superconductive tunable filters.

Recently, microwave devices of the high-temperature superconducting thin film operating in a high magnetic field have been studied. In such applications, it is necessary to create a thin film with a low microwave surface resistance (Rs) under a high magnetic field. The introduction of artificial pins (Aps) in YBa2Cu3O7 (YBCO) thin film is useful to reduce the Rs in high magnetic fields. We examined the formation of APS by ion implantation of Si, Mo and In ions. We found that Rs of ion-implanted YBCO thin films could be reduced to about one-quarter of the non-irradiated YBCO thin films. We also examined the APs structure by positron annihilation life time spectroscopy. As a result APs was found to be a vacancy of about 0.5nm3. We developed a prototype 700MHzNMR pickup coil made by YBCO thin films. The sensitivity of NMR could be improved by using YBCO pickup coil compared with using a conventional NMR pickup coil. It is expected to improve further sensitivity by using the ion-implanted YBCO thin film, and we are currently considering it.

Quantum information processing devices are now sufficiently advanced to imagine scaling up this technology into complex, multicomponent systems. This transition, from quantum devices to quantum machines will require the development of new cryogenic electronic platforms operating at 4 kelvin and below. This talk will outline efforts to develop cryo-CMOS and SiGe circuits for controlling and reading out scaled-up quantum machines.

Flexible and robust superconducting microwave transmission line cables are an enabling technology for densely integrated cryogenic electronics systems. Currently, single co-axial transmission lines are bulky and often limit integration density due to volume and thermal load constraints. We will discuss the current status of our research and development efforts to construct and characterize multi-conductor superconducting flexible cables made using thin-film processing techniques. Trade-offs and performance of various materials stack-ups will be discussed. Characterization has been performed using transmission line and resonator geometries from temperatures of approximately 9 K down to 20 mK. These cables show great promise as interconnect structures in future superconducting and cryogenic electronics systems, including superconducting quantum computing applications.

In the era of internet of things (IoT) and wireless communication, energy- and area-efficient wireless transceivers are critical for extended battery life and small form factor of many systems. On the other hand, innovations in analog circuits have been driven by rapidly evolving semiconductor technology in line with Moore’s law. A switched capacitor power amplification (SCPA) technique, based on RF switched capacitor digital-to-analog converter architecture, offers very high energy efficiency and superior linearity for wireless transmitters. The measured peak Pout and power-added-efficiency (PAE) of a prototype SCPA are 25.2 dBm and 45%, respectively. For 802.11g 64-QAM OFDM modulated signal, the average Pout and PAE are 17.7 dBm and 27%, respectively, and the measured EVM is 2.6%. Class-G technique can be applied to SCPA to further enhance the efficiency. With additional supply voltage and advanced switching scheme for multiple supply voltages, the measured peak Pout and PAE are 24.3 dBm and 43.5%, respectively, whereas the average Pout and PAE are 16.8 dBm and 33%, respectively, for 802.11g signals. The measured EVM is 2.9% without any predistortion applied.

Recently, we have developed a transmitter based on RFDAC. A three-level LO clock is employed to improve the digitizing efficiency and to combine I and Q signals in time domain. Since the signals are in a single stream, I and Q signals could be up-converted using the one sampling mixer, realizing I/Q sharing structure. The mixer thermo-code element pairs whose output powers are internally cancelled, are disabled to reduce the power consumption. The resulting RFDAC provides the efficiency comparable to the polar transmitter without using a Cordic. The output powers of the mixer elements are amplified by inverters and the final powers of the cells are combined using a capacitor array similarly to the switched capacitor DAC. To reduce the thermo-code elements, the data is processed only in the first quadrant and is rotated to the original position. A dual Vdd structure is also employed. The detailed design issues and the measured performances of the transmitter will be discussed.

In this talk the switched capacitor PA (SCPA) is introduced. It leverages CMOS inherent strengths of fast switching and lithographic matching to yield a linear, efficient digital PA. The original SCPA was a polar PA, subject to significant system level non-linearity (wide bandwidth, lack of synchronization, etc). I will introduce several techniques that implement SCPAs in discrete phase spaces; several multiple phase digital PA architectures will be discussed that alleviate the need for wideband phase modulators and synchronization. I will highlight several recent examples from the University of Utah PERFIC lab’s research with applications of the multiphase techniques to the SCPA.

The recent trend of deploying digital transmitters has stimulated an increasing interest in power amplifiers research and development using RF power DACs. The digitally intensive and reprogrammable nature of RF power DACs opens the door to creating hybrid and broadband power amplifier architectures that can combine the advantages of different power amplifier techniques and potentially achieve superior peak power efficiency, back-off efficiency, and linearity. These hybrid architectures cannot be easily realized using conventional analog power amplifiers. Two power amplifier design will be presented as examples of such hybrid digitally intensive power amplifiers. One design exploits the real-time cooperation of dynamic load modulation and class-G supply modulation, and the other design shows a power amplifier with broadband and highly linear operation.

Analog/RF circuits, particularly conventional power amplifiers, do not benefit much from technology scaling. RF transmitters utilize multiple passive components which do not scale with technology and are narrowband. Hence an architecture that can benefit truly from technology scaling as well as offer wideband multi-mode performance will be beneficial. Such architecture is indispensable for reconfigurable or software-defined radios. Digitally modulated transmitters offer such a solution. In this talk, we will cover the pros and cons of a digitally modulated transmitter with special emphasis on a power amplifier. Ways of achieving higher efficiency by using novel impedance modulation will be introduced. In particular, a high power 65nm mixed-signal Inverse Class-D power amplifier design with switchable power combiner will be presented and its integration into a complete transmitter will be discussed.

The insatiable need of consumers worldwide for higher data rates in wireless communication brings the frequencies of operation towards the millimeter wave spectrum. The polar architecture enables the power amplifier to operate in saturation where efficiency is highest, even when handling higher-order modulations with variable envelope. System analysis for IEEE 802.11ad applications show a required input baseband signal bandwidth of more than 1GHz and the need to synchronize the amplitude and phase paths with picosecond time resolution. A prototype chip uses an RF-DAC with 10GS/s to modulate the amplitude path while overcoming out-of-band spectral images. Implemented in 40nm bulk-CMOS, the 0.18mm2 core circuit transmits a full-rate 5.3dBm QPSK signal with 15.3% average PA efficiency and -23.6dB EVM with -30dB out-of-band distortion.

Pulse-width modulation (PWM) encodes the amplitude information of a signal in the duty cycle of a periodic pulse waveform that is switching at a frequency much higher than the bandwidth of the signal itself. Combined with a switch-mode driver amplifier, such as a class-D stage, PWM offers an efficient approach for implementing transmitters. The approach is especially well-suited for digitally-intensive, CMOS-friendly implementations. While PWM combined with a class-D output stage has been very effectively employed in audio systems, its use in wider bandwidth applications such as wireless systems requires an exploration of new circuit techniques and architectures. In this talk, we will describe PWM-based transmitter architectures for applications where the signal bandwidth can extend from several hundreds of KHz to tens of MHz. Transmitters that allow for PWM generation at RF, without the requirement for frequency upconversion, will be described. An overview of the benefits and potential limitations of the PWM approach will be presented. Practical implementations of wireless transmitters employing this approach will also be presented.

The first published THz medical imaging appeared in the literature at the end of the 1990s and the field experienced accelerated research and development well into the 2000s. The vast majority of reported systems were THz time domain spectroscopy (TDS) and THz time domain imaging (TDI) where ultra-broadband pulses with typically > 1 THz of instantaneous bandwidth were generated and synchronously detected with some combination of photoconductive and electro-optic devices driven by femtosecond lasers. Fast forward to today and while much progress has been made in clinically relevant investigations, clinical translation of THz technology has been limited. There are many applications where TDS and/or TDI still seem ideally suited. However, there is growing evidence that THz imaging systems based on RF technology (part of the standard tool set in the fields of personnel imaging, non-destructive evaluation, and radio astronomy) may be well matched for a multitude of clinical investigations. These issues strongly suggest broad instantaneous bandwidth may not be necessary and thus that the field of THz medical imaging may benefit greatly from research and development of systems based on RF technology. In this talk we introduce two medical applications where video rate acquisition of large areas of the body in pursuit of high contrast anomaly detection would immediately generate interest from the medical community. Further, physiologic and clinical work flow arguments will be made that highlight some of the limitations of TDS/TDI and emphasize the unique capabilities of RF based systems.

Electromagnetic waves in the millimeter- and sub-millimeter-wave frequency ranges are used in fast-scan rotational spectroscopy to detect gas molecules and measure their concentrations. This technique can be used for indoor air quality monitoring, detection of toxic gas leaks, breath analyses for monitoring bodily conditions and many others. This talk reports a 210-to-305GHz transmitter and receiver circuits for a rotational spectrometer. Techniques presented include on-chip antennas, artificial magnetic conductors, diode-based mixers, and high-precision PLLs. Full spectroscopy measurements using these CMOS components will also be reported. 200-250-GHz transmitter and receiver for rotational spectroscopy, which can identify a wide variety of molecules in gas phase and quantify their concentration are demonstrated in 65-nm CMOS. Because the width of spectral lines is ~1MHz or Q is more than 200,000, lines of different molecules do not overlap and provide almost absolute specificity. A rotational spectrometer is an electronic nose. Considering how smell is used in daily life, the number of applications should be almost limitless. The transmitter and receiver were used to detect ethanol from a human breath demonstrating that CMOS will be able to support practical applications at 200-300GHz

The talk will present the latest work at UCSD for wideband imaging systems at 10-40 GHz. Transmitter and receiver chips, achieving record performance, will be presented. Also, a simple imaging system showing the resolution of this technique will be shown.

This talk will discuss the challenges of LO requirements and LO distribution for radar and imaging arrays specifically looking at phase noise and distortion related effects and show examples of physically disconnected phase array exciters at 100 and 150 GHz in CMOS technology

Recently, terahertz time-domain spectroscopy (THz-TDS) and terahertz imaging are new methods with unique capabilities. Based on coherent and time-resolved detection of the electric field of ultrashort radiation pulses in the far-infrared, and very versatile ways to generate and detect, Terahertz and millimeter waves can penetrate various dielectric materials, including plastics, ceramics, crystals, and concrete, allowing terahertz transmission and reflection images or analyses to be considered. In this review, the authors describe the techniques in its various implementations for static and time-resolved spectroscopy and imaging, and illustrate the performance of the technique with recent examples with various laboratory and industrial projects. Possible future improvements are related to semiconductor or optical laser sources and detector (Quantum Cascade laser and MMIC sources and sensors, THz camera for example). The terahertz science and especially terahertz imaging will be probably an emerging and an efficient tool for a lot of industrial applications.

The performance of a differential LC oscillator can be enhanced by resonating the common-mode of the circuit at twice the oscillation frequency. When this technique is correctly employed, Q-degradation due to the triode operation of the differential pair is eliminated and flicker noise is nulled. Although the original topology using this technique first appeared in 2001, its performance remains state of the art. In this workshop, we will use Bank’s general result to show that common-mode resonate topologies are, in fact, near-optimal. That is, for a given resonator tank Q, the FOM of such a design is within 1dB of theoretical limit for any oscillator. Common design pitfalls, which may prevent this near optimal behavior, will be explained and the importance of differential pair sizing will be explored. Insights from class-D and class-C topologies will be used to gain additional intuition into the optimization of the differential pair. It will also be shown how recently published topologies achieve common-mode resonance using a single differential LC tank (i.e. removing the requirement for an additional tail inductor). Although such topologies have one less degree of design freedom, their performance is theoretical identically to the standard approach of using a tail resonate tank.

Spectral purity of RF LC-tank oscillators is typically addressed by enhancing its oscilla-tion voltage, improving the quality factor of the tank, lowering its noise factor, and in-creasing its power consumption. However, the rate of improvement in phase noise versus power consumption reduces as the core devices enter the triode region. On the other hand, technology scaling limits the oscillation voltage swing due to the reliability issues. It also slightly degrades the tank Q-factor and transistor excess noise factor and thus penalizing oscillator phase noise. Conse-quently, the oscillators of excellent spectral purity and power efficiency are becoming more and more challenging. This has motivated an intensive research leading to recently introduced new oscillator topologies. In the first part of this presentation, we specifically address the ultra-low phase noise design space while maintaining high power efficiency. The main idea is to enforce a pseudo-square or clipped voltage waveform around the LC-tank by increasing the second or/and third harmonics of the fundamental oscillation volt-age through additional impedance peaks, thus giving rise to a class-F operation. As a result, the oscillator impulse sensitivity function and circuit-to-phase noise conver-sion reduce especially when the active gm-devices periodically enter the triode region during which the LC-tank is heavily loaded. In the second part of this talk, we switch gears to the ultra-low voltage and power design space for Internet-of-Things (IoT) applications. The benefits and constraints of different flavors of LC oscillators are investigated from this perspective. It will be shown that a switching current-source oscillator combines advantages of low supply voltage of the conventional NMOS cross-coupled oscillator with high current efficiency of the comple-mentary push-pull oscillator to reduce the oscillator supply voltage and dissipated power further than practically possible in the traditional oscillators.

The talk will deal with the design of low-phase noise voltage-controlled harmonic oscillators (VCOs) implemented in bipolar technologies. The design challenges related to achieving minimum phase noise for a given set of technology parameters (supply voltage, metal stack, varactor devices, etc.) will be discussed with particular emphasis to attaining low phase noise while using varactor diodes, to the use of magnetic transformers in the resonator, and to the selection of the most appropriate oscillator topology. Suitable design techniques to tackle such issues will be illustrated. As a design example of the use of the proposed techniques, two class-C VCOs tailored for operation in the K-band will be presented.

Modern mobile communication systems need clocks with very low phase noise and low power consumption. In RF and mm-wave oscillators this can be achieved acting on the oscillator topology and/or on the resonator quality factor. Oscillator topology affects the power vs phase noise trade-off in two equally important ways. First, acting on the conversion of circuit noise into phase noise through changing the impulse sensitivity function (ISF); second, changing the maximum achievable power conversion efficiency, i.e. the conversion of DC power into resonator RF power, which directly affects the phase noise. The goal of this talk is to investigate the ultimate performance limit for some of the most used oscillators topologies, including most types of class-B (standard, AC-coupled and with tail filter), class-C and class-F LC oscillators as well as mm-wave distributed oscillators (traveling-wave and standing-wave). An intuitive yet sufficiently accurate formulation of phase noise is presented. To compare different topologies an excess noise factor that represents the difference between the maximum achievable Figure of Merit and the actual one is also introduced. In addition, the theory is experimentally verified in a rigorous and objective way comparing different topologies in the exact same operating conditions, i.e. technology, Q of the tank, dividers, etc. Measurements on several chip prototypes allow to verify, in an unbiased way a very good agreement between the model and both simulations and measurements.

Low frequency LC-VCOs have been well studied in the literature and several strategies have been developed to optimize their performance. However, several interesting challenges in the mm-wave space, especially close to the fT/fmax, motivate the need for a closer examination of the tuning range and phase noise in mm-wave VCOs. This seminar will elucidate some of the key challenges in designing mm-wave VCOs while also offering some insight on how to design them “robustly” in silicon-based technologies. First a detailed analysis of the ultimate performance bounds in simultaneously achieving low phase noise and wide tuning range will be presented. Next, the impact of technology scaling on the achievable performance bounds will be illustrated. The tuning range and close-in/far-out phase noise performance of MOS and HBT VCOs across the 10-70 GHz space will then be analyzed and compared. Here, special attention will be paid to the impact of different circuit parameters on the phase noise performance in both VCOs. Finally, a new BiCMOS topology is elicited and demonstrated to meet stringent phase noise and turning range specifications.

Advanced InP HEMT Technology for Terahertz Amplifier Circuits Advances in scaled InP High Electron Mobility Transistor (HEMT) processes have achieved record low noise amplifier performance from millimeter-wave to THz frequencies and have resulted in the first ever Terahertz Monolithic Integrated Circuit (TMIC) demonstrating amplification at 1.0 THz (1000 GHz) for the first time. This presentation will describe the key manufacturing advances and future technology development roadmap, along with challenges as we transition advanced InP HEMT nodes from R&D into fielded hardware.

This presentation will compare the high frequency performance scaling of SiGe HBTs and MOSFETs to 2-3nm gate length and beyond 2THz transistor fMAX based on technology CAD (TCAD) and atomistic simulations. Characterization techniques and S-parameter measurements of state-of-the-art silicon MOSFETs, SiGe HBTs, and of a variety of HBT-HBT and MOS-HBT cascodes from DC to 325 GHz will be discussed along with simulations of the scaling of analog and mixed-signal mm-wave benchmark circuit performance from the current to future technology nodes.

The talk will present the latest development in THz phased-array transmitters, with emphasis on operation frequencies greater than 200 GHz. This includes CMOS phased-arrays operating at 370-410 GHz with high efficiency on-chip antennas and record EIRP, high efficiency multipliers and multiplier-arrays at 200-240 GHz, and quadruplers at 400-500 GHz with two-dimensional power combining.

Silicon-based integrated circuit technology provides a great platform for enabling compact, efficient, low-power, chip-scale THz systems for new applications in sensing, imaging and communication beyond the niche scientific applications that the spectrum is currently known for. While this is partially facilitated by scaling that has pushed device cut-off frequencies (ft, fmax) up into the higher mm-Wave and THz frequency range, the true paradigm shift in silicon integration is that it provides a unique opportunity to enable a new field of active THz electromagnetics realizable through a circuits-EM-systems codesign approach. At these frequencies, the chip dimension is several times larger than the THz wavelengths which allows novel scattering and radiating properties in a substrate that simultaneously supports a billion high-frequency transistors that can generate, process and sense these signals. The ability to actively synthesize, manipulate and sense THz EM fields at subwavelength scales with circuits opens up a new design space for THz electronics. THz architectures emerging from this space are often multi-functional, reconfigurable and break many of the classical trade-offs of a partitioned design approach. This talk will provide examples to illustrate this design methodology on THz signal generation with beam-forming and spectrum control and THz spectrum sensing.

Terahertz circuits, interconnects, and packaging have received unprecedented attention in recent years for their use in emerging areas such as security screening and standoff weapons detection, high speed digital communications, automatic landing systems through fog and dust, and even for aircraft re-fueling in air. Traditional terahertz application areas such as high-resolution spectrometers and imagers for astrophysics, planetary, and Earth science instruments are increasingly looking for multi-pixel array architecture which requires high level of integration and packaging. Active components such as InP high electron mobility transistors (HEMT) and metamorhpic HEMT (mHEMT), heterojunction bipolar transistors (HBT), GaAs Schottky diodes, and CMOS circuits are being used at these frequencies. However, one major challenge has been to integrate them in a multipixel detector system with low-loss interconnects and build a highly integrated package. Common interconnect circuits such as microstrips and CPW lines are too lossy at these frequencies. Moreover, conventional approach of building single-pixel receivers and stacking them to assemble multi-pixel array receivers is not suited at terahertz frequencies. What one needs are novel ultra-compact receiver architectures which are easy to fabricate, preferably by lithographic techniques, to build multi-pixel array receivers where majority of the front-end components along with the antenna element can be integrated in a small form factor. In this workshop presentation we’ll talk about different multi-pixel receiver architectures at terahertz frequencies, specifically focusing on silicon micro-machined front-end components. We’ll discuss novel stacking of micro-machined silicon wafers which allows for the 3-dimensional integration of various terahertz receiver components in extremely small packages which easily leads to the development of 2-dimensioanl multi-pixel receiver front-ends in the terahertz frequency range. Integrating antennas with terahertz front-end elements has been challenging as most of the planar antennas are too lossy at these frequencies. In this presentation we’ll also explore novel horn and lens antenna architectures which are suitable for integration with silicon micromachined receiver systems. It will be shown that using advanced semiconductor nanofabrication techniques it is possible to design, fabricate, and demonstrate a super-compact, low-mass, and highly integrated multi-pixel terahertz array receiver. The research described herein was carried out at the Jet Propulsion Laboratory, California Institute of Technology, Pasadena, California, USA, under contract with National Aeronautics and Space Administration.

An active research and development of terahertz (THz) technologies were initiated with use of photonics technologies in 1990’s. In the last 10 years, with the advance of semiconductor devices and integrated circuits, electronics-based THz technologies have recently gained a great attention to make THz systems and subsystems more compact and cost-effective. In this talk, we review photonics-based approaches in practical applications such as communications and measurements, clarify merit, role and issue of photonics in THz technologies, and discuss its future directions to compete and/or coexist with electronics.

Scaling of silicon technology following Moores’ law has allowed feasibility of CMOS circuits operating above 100GHz. These high frequencies support order of magnitude higher bandwidths enabling high-data rate applications. Furthermore, at these high frequencies, thin (mm-range) polymer fibers such as those made from PE,PP, PS, PTFE, are excellent transmission media and exhibit fairly low loss, below 5dB/m. As such, the combination of CMOS mm-wave transceivers, on-chip or on-board antennas and thin plastic fibers leads to an innovative communication concept that is, in certain applications, perform better than optical communication or copper wireline communication. Especially for cases where high EMI resilience, high mechanical tolerance and low costs are important, such as automotive communication, this 'RF over Plastics' concept can be a game-changing technology. This presentation will discuss some of the key benefits and drawbacks of polymer microwave fiber technology and will present the results of the ongoing research at KU Leuven on this topic since 2012.

Integrated circuits in silicon technologies have been increasingly used in the THz spectrum for synthesis and detection of signals. In spite of the physical limitations related to device speed and cutoff frequencies, circuit design techniques have been implemented to enable integrated terahertz systems well beyond these frequencies. THz imaging in silicon is example of this approach, and can potentially enable new applications in this range. The large available bandwidth from 0.3-3THz combined with on-chip multi-element integration can enable novel imaging perspectives. However, enabling this system in a cost-effective and scalable fashion presents significant challenges not only in the realm of circuit design, but also in the co-design of the entire system. Many challenges are still to be addressed, starting with availability of efficient sources with high radiated power, improved tuning range and coherence. On the receiver front, detectors are also in need of improving their sensitivity. These considerations have led to active imaging approaches and systems have been demonstrated with these methodologies. This workshop presentation will focus on recent work that addresses these challenges in silicon technology. This will cover circuit realizations in CMOS technology for terahertz generation and detection, as well as novel imaging approaches.

This presentation will give an overview of 5G communication Systems including New Ecosystems and Markets, Usage Scenarios, Mission-Critical Services, Massive Internet of Things, Standards and Spectrum. Major trends will be summarized with a perspective on developments and technologies at this extraordinary time in communications. These important areas include: Open agile innovation, Cloud based applications with wireless connectivity to devices, Local/distributed integrated smart capabilities (processing and memory), and Increasing Software-defined-centric world The development of introduction of past 2G, 3G and 4G telecom systems were done in a well-defined and established ecosystems but NOT in 5G, it will be mix of established players and new ecosystems. Examples of affected ecosystems will be given beyond the world the Internet Giants of Google, Facebook, Amazon, plus others unexpected etc.

Recent years have seen considerable increase in activities related to 5G definition and associated research. While a number of novel use cases and services are being contemplated, there is consensus that data demand will continue to grow, putting further pressure on already congested spectrum. Millimeter-wave bands have been cited as having the potential to alleviate some of this pressure. This talk will cover the opportunities and challenges with mobile communications in the millimeter wave band for the deployments and use cases of interest. Specifically, the talk will first cover material and channel propagation measurements highlighting the contrast with sub-6GHz propagation. With those observations, a number of system design principles and associated device/component level requirements will be outlined. Finally, taking some of the design principles and component level considerations into account, coverage and capacity modeling results will be presented. Some test results from our first generation prototype system at 28GHz will also be presented

Millimeter-wave (mm-Wave) technology has matured and it’s almost ready for 5G. However, the limited capacity of back-haul links are becoming bottleneck of mm-Wave integrated networks. In this talk, a fusion of mm-Wave access and mobile edge computing (MiEdge) is introduced to alleviate the problem of back-haul links and to facilitate spreading of mm-Wave technologies in 5G systems.

The exploration of mm-wave frequencies and active antenna architectures will pose completely new challenges in the design of base stations for 5G mobile networks. In the first part of this presentation, we will discuss how various RF hardware impairments will affect the performance in emerging base station architectures. We will introduce new analysis techniques that help us explain and understand how key RF parameters like nonlinear distortion, energy efficiency, and phase-noise will have to be revisited in derivation of future circuit design requirements. The theoretical predictions will be complemented with experimental results obtained using Chalmers 30 GHz massive MIMO test-bed. The second part of the presentation will be devoted to the design of critical mm-wave circuits for 5G base stations. Starting from 5G base station requirements, we will present an integrated 30 GHz SiGe BiCMOS transmitter lineup comprising the co-design of analog pre-distortion linearization, I/Q modulator, and efficient Doherty PA circuits. The excellent results obtained demonstrate the capabilities offered by advanced semiconductor technologies, and indicates a possible direction towards realization of efficient and linear transmitters for next generation wireless systems.

Traditionally, in a wireless communication system, the power amplifier (PA) has been designed as a standalone block. More recently, the need to maximize the performance and reduce the cost lead the PA and digital pre-distortion (DPD) algorithms to a joint development. For the future, besides the referred performance maximization and cost reduction, the 5G systems will add an extra parameter, the integration. Considering that the massive MIMO antenna structures (for frequencies below 6GHz) might contain 64 or more antenna elements (each one with an associated PA), it is evident that to achieve the requested objectives, the PA design cannot be dissociated from the remaining front-end blocks. This factor is even more visible at mm-Wave frequencies, at which the physical dimensions are so stringent that integration is even more important and the technology limitations forces more innovative designs.

This presentation will discuss the challenges of implementing a wide-band and high-efficiency PA in CMOS technology at mm-wave frequencies. Key-enabling circuit techniques, such as on-chip power combining and wide-band AM-PM cancellation, will be discussed in greater detail. These techniques are crucial to support higher order constellations such as 64QAM and 256QAM as needed for 5G. Several PA examples in 40nm and 28nm CMOS, operating at 28GHz, 60GHz and 85GHz will be used throughout the presentation to further explain the used techniques.

The talk will present the latest development in 5G communication systems at UCSD. The phased-arrays and related communication links will be discussed, both at 28 GHz and at 60 GHz. Prof. Rebeiz group has achieved Gbps over hundreds of meters, even kms, using phased-array technologies. All work is based on silicon RFICs and innovative packaging

As the demand for higher data rates in mobile communications still increases, new wireless systems might even use millimeter wave frequencies for the connection between the base station and the mobile. In addition, millimeter wave links are very attractive for front- and back-hauling of the new femto-cell base stations placed in areas of very high user density like transportation hubs and downtown areas. Since at millimeter wave frequencies they use of cables between antenna and transceiver is not anymore reasonable in this presentation antenna concepts for the above sketched applications will be presented together with suitable integration concepts.

The first decade of millimeter-waves in silicon (2000-2010) saw the maturation of complex SiGe and CMOS integrated circuits and systems for short-range high-data-rate wireless communications. More recently, the millimeter-wave range has drawn significant interest for next-generation ("5G") cellular communication networks. Such applications place significantly different requirements on the systems, including link range, mobility and coexistence, resulting in far more stringent circuit requirements, such as higher transmitter output power, stricter transmitter linearity, and agile beam-steering in large-scale phased-arrays. In addition, advanced communication paradigms, such as full-duplex and massive MIMO, are being considered to further enhance the spectral efficiency and data capacity. This workshop presentation will reviews recent developments on a few candidate CMOS-based circuit and system technologies for 5G millimeter-wave applications.

Next-generation wireless communications systems call for RF transceivers with multi-functional operability as well as immunity to undesired interference and noise that are typically present in dense communication environments. As a result, advanced filtering devices capable of meeting the very stringent requirements demanded by these systems need to be conceived. In particular, acoustic-wave (AW) filters have been the key filtering technology of mobile transceivers due to their high quality factor (>10,000) and miniaturized volume (<4 mm3). However, due to their ultra-narrow fractional bandwidth, limited type of realizable transfer function (band pass type), and lack of reconfiguration, the scaling of RF front-ends to multiple bands and wide-band signal processing remains as a great challenge. Within the scope of this workshop, new hybrid-lumped-element-AW-based filter developments that allow to overcoming the aforementioned shortcomings in terms of bandwidth, multi-band operation, and transfer-function adaptiveness will be presented. Moreover, RF interference-mitigation techniques for ultra-wideband wireless communications transceivers based on new broad-band filtering components with fully-adaptable in-band notches will be described. It is believed that these filter design techniques will be essential for upcoming wireless communications systems including 5G.

Coping with the rapid increase of mobile data traffic, the frequency spectrum for mobile communications has been expanding, and in the 5G system a wide range from sub-6GHz to mm-wave bands will be employed in addition to conventional LTE bands. For mm-wave communications, beam-forming technique and/or Massive MIMO are indispensable technologies, so new requirements for RF front-ends have been emerging for both mobile terminals and base stations. This presentation describes the challenges for higher and wider frequency range application of RF front-end devices and modules. For a 28GHz phased array unit, Antenna integrated Module (AiM) with small size and high performance modules including not only RF-ICs, amplifiers and passive components but also antenna pattern using a low temperature co-fired ceramics (LTCC) is one of the key technologies. Heterogeneous integration and 3D structure lead to the size reduction and low loss. Furthermore, a quartz waveguide Band Pass Filter (BPF) and a miniaturized circulator for 28GHz improve the massive MIMO antenna system. Expansion to 60GHz antenna array integrated modules and sub-6GHz reconfigurable RF front-end architectures are discussed. Accurate and high speed measurement technologies for mm-wave antenna module performances are also introduced.

The talk will present the latest development in 5G communication systems at UCSD. The phased-arrays and related communication links will be discussed, both at 28 GHz and at 60 GHz. Prof. Rebeiz group has achieved Gbps over hundreds of meters, even kms, using phased-array technologies. All work is based on silicon RFICs and innovative packaging.

Small cell backhaul communication in 5G mm-wave systems will occur in the 27 – 30 GHz band. Depending on the choice of antenna plane, link budget and beam steering concept, the required transmitted output power of the Front End Modules (FEM) is in the range of 15dBm to 30dBm. Besides the challenge to deliver this level of power at mm-wave frequencies in silicon technology, efficiency becomes a crucial design parameter as case cooling of the antenna panel with more than 100 FEM units is preferred. In this presentation we will discuss techniques to realize silicon-based RF PAs for these power levels at these mm-wave frequencies and elaborate on thermal design and packaging. In this workshop we will discuss a 5G mm-wave transmit line up with 18dBm output power and an integrated 30dBm PA for backhaul communication, both realized in an in-house SiGe:C BiCMOS technology.

The last 15 years has witnessed revolutionary changes in mobile computing and wireless communication. This was fueled in large part through Moore’s Law, coupled with research and development of new highly-integrated, silicon CMOS devices which transformed large bulky transceiver components into a single chip for wireless applications. These single-chip radios freed up valuable space for more memory and powerful processors, making the modern smartphone, as we know it today, so common and ubiquitous. Although the architectures, circuits, and system-level design methodologies to realize these low-cost, highly-integrated RF ICs have largely been defined, questions remain on how to enable chips for emerging applications in an era of large scale data acquisition, and communication, for a variety of devices ranging in use from wireless sensing, to high-speed mobile communication and radar. A common theme among these future devices is the need for highly-integrated ultra-broadband (10GHz+) transceiver solutions. However, achieving high bandwidth in the mm-Wave band becomes challenging without burning excessive power and occupying an absolute minimum silicon area, particularly in phased-array applications where numerous elements are replicated on the same die. This presentation explores recent work which attempt to achieve extremely high bandwidth transceivers while utilizing the smallest silicon area possible.

A major challenge for low-cost silicon-based mm-wave wireless systems, e.g., the 5G MIMO communication links, is to provide large transmitter (Tx) output power (Pout) with high energy efficiency and linearity from a limited supply voltage, so that the high path loss and limited link budget at mm-wave can be compensated. Power combining is often required for these mm-wave Tx. The existing power combining techniques are mainly in two categories. Passive networks can combine the Pout from multiple power amplifiers (PAs) and feed the single antenna port. However, lossy power combiners and large impedance transformation ratios degrade the total Pout delivered to the antenna and lower the Tx efficiency. Alternatively, spatial power combining using antenna array increases the total EIRP but at the expense of a large array panel size. Moreover, a large antenna array often presents an exceedingly narrow (or even pencil-sharp) beam-width; this complicates the Tx/Rx alignment and is challenging for dynamic and mobile mm-wave applications, such as 5G links. In addition, adding silicon lens enhances EIRP but increases cost and packaging complexity. In this talk, we present a concept of multi-feed antenna (MFA) and its co-design with mm-wave transmitters; MFA can be viewed as multiple electrically small antennas that are driven concurrently by multiple feeds but radiate collectively and efficiently as a single antenna. Such an MFA structure naturally allows on-antenna low-loss power combining from multiple PAs, radiation impedance down-scaling, and boosting total output power in one single antenna footprint. We will present multiple MFA designs at different frequencies to demonstrate the concept. We will also present a 60GHz linear radiator element in a 45nm CMOS SOI, as an on-chip MFA driven by 16 linear PAs. The radiator IC generates 27.9dBm Psat and 33.1dBm peak EIRP with 23.4% PAE at 59GHz, showing the best reported mm-Wave PA/transmitter performance in the 60GHz band. Without any signal pre-distortion, the radiator IC achieves -21.9dB EVM with 20.2dBm Pavg for 4Gb/s 16QAM signal, and -25.4dB EVM with 19.3dBm Pavg for 4.8Gb/s 64QAM signal.

Phased array mmWave transceiver architectures for directional Gb/s wireless links, suitable for full integration in silicon are reviewed. Design considerations of building blocks, key to beamforming performance such as RF phase shifters and variable gain amplifiers are presented and illustrated with examples. Antenna diversity and beam forming techniques suitable for low-cost packaging implementation are also discussed. Fully integrated transceiver implementation examples including antennas-in-package are presented to illustrate system-implementation trade-offs, and IC-package co-design challenges.

Traditionally antennas, packages, RFICs are designed separately and put together with interconnects and matching networks to create wireless communication systems. Recent advancements of CMOS-based ICs are enabling mm-wave systems for next generation communication. At mm-waves, it is desired that the antennas, mmW-ICs, packages, system-modules are co-designed to avoid performance degradation. This talk will present mm-wave co-design and testing examples to discuss the advantages and challenges associated with CMOS-based mmW systems for 5G applications.

The increase of integrated radio transceivers carrier frequencies is posing an important challenge for testing. 5G system are following in this directions with envisaged carrier frequencies beyond 24 GHz. Built-in testing and built-in self-testing (BIST) for RFIC is an attractive solution for RFICs but with increasing carrier frequencies the access to the high frequency outputs of the circuits, especially at mmW frequencies becomes very challenging if a minimum impact on the circuit performance has to be preserved. Non-invasive, contact-less techniques using indirect measurements, such as the local temperature increase have been recently proposed to tackle this issue. Other recent BIST techniques are based on replica circuits. We will introduce in this talks these innovative BIST technique for mmW integrated front-ends.

A number of new design and test challenges arise as 5G evolves to provide greater data capacity. Modulation bandwidths much wider than LTE are expected and designs are done at frequencies above 28 GHz where spectrum is available. These millimeter wave devices may not have a physical test port, requiring test equipment to connect wirelessly and calibration methods that correct for the air interface. This presentation introduces challenges associated with Physical Layer (PHY) testing of SoC at Millimeter Wave frequencies; then it discusses generation and analysis tools offered by Test and Measurement vendors to improve signal quality measurements at these frequencies. Results achieved on a 60GHz low-power transceiver module are presented.

To achieve the dramatic improvements in capacity and spectral efficiency needed to accommodate access to high volume of wireless data and the even increasing number of users who want to access to it anytime and anywhere, three symbiotic technological directions are independently emerged: (1) A push towards greater frequency reuse through the creation of smaller and smaller cells, referred to as pico- and femto cells with ranges on the order of 10-200 meters. (2) A consideration of millimeter wave frequencies around and above 70 GHz where the spectrum is less crowded and greater bandwidth is available. (3) The idea of base stations equipped with a large number of antennas that can simultaneously accommodate many co-channel users. This idea is referred as MIMO. However, the next generation multi-antenna transceivers must also provide sufficient output power with great radiation selectivity and directivity. This notion is calling for new generations of MIMO transceivers that also provide beamforming forming and spatial power combining. This talk will be going through latest advances in this exciting domain.

Design by Mathematics is an inventive design approach dedicated to high performance integrated circuits. It is based on mathematical principles and techniques, such as Riemann’s integration or Fourier’s transformation. These mathematical tools are used to optimize a specific signal processing and conditioning. A given tool behavior is then copied as much as possible within a silicon implementation, yielding to mixed-signal integrated circuits that demonstrate innovative system architectures and disruptive approaches. While using Design by Mathematics does not imply one will achieve better performances than when using classical design techniques, it offers a substitute that can counteract key technical bottlenecks and pave the way to new opportunities. In this talk several Design by Mathematics examples will be presented, focusing on wireless systems. These systems include next generation standards such as 5G and its carrier aggregation technique in the radio frequency range. Fourier’s and Walsh’s transformations will used, as well as Fourier’s recombination and Riemann’s integration, for either the receiver path or the transmitter path of a system.

Emerging low power radios are going to be required for IoT, WBAN and even 5G. In this tutorial we will explore the power dissipation limiters of current radio architecture and propose solutions to counteract them. In particular, we will describe details for a standard compliant 2.4GHz low power radio architecture for IEEE 802.15.16 WBAN radios that uses both a new receiver architecture and a new transmitter architecture to approach the single milliwatt power level while being standard compliant.

tbd

A novel single-PLL, fixed-oscillator, wideband multi-tuner architecture is presented for concurrent multi-band and multi-channel car radio reception based on wideband and high-resolution A/D converters. This architecture simplifies the LO and clock generation, while it also prevents oscillator pulling and spurs which are notorious for classical narrow-band multi-tuner solutions. An implementation of a wide-band HD radio & DAB/T-DMB receiver will be shown, demonstrating best-in-class blocker performance (DAB FoS up to 70dBc) in combination with state-of-the-art DAB sensitivity down to -102dBm. The design aspects of a high-performance 2.2GHz DS ADC with -102dBc THD are presented that also enables the wideband fixed-oscillator receiver concept for the very performance demanding automotive AM/FM radio standards.

tbd

Transceiver architectures and circuits for the 5G technologies will be reviewed in this presentation. Gb/s data-rates over the air could be achieved in the sub-6GHz cellular bands as well as sub-60GHz and sub-30GHz mmWave bands. 5G technologies that support data-rates in excess of 1GB/s could be implemented in any of the existing bands as long as they yield. RF front-end, transceiver, and modem architectures that could offer the highest (by humans noticeable) data-rate in handsets would be the obvious winner, no matter what RF frequency they ‘pick’ from the air. In this presentation, the sub-6GHz cellular transceiver architectures would be compared to the sub-60GHz and sub-30GHz mmWave transceiver architectures in terms of performance, area, power consumption, and cost. Could the big bandwidth of the mmWave design win over the high-order modulation schemes of the well-established cellular design? Or would the cellular and mmWave transceivers have to co-exist and co-work together for the best of both worlds? These are just two of the questions that will be addressed in this presentation.

Gigabit Wireless LAN is here! Sustained, multi-client, Gbp/s WLAN is entering wide-spread deployment in enterprise networks. WLAN is on the cusp of being able to replace wired Ethernet as we know it, not only in consumer application, but also in enterprise and industrial applications. This talk will discuss the key techniques that enables Gbp/s WLAN transceiver, including digital-intensive circuit architectures in PLL and transmitter design. Built-in, self-calibration necessary to obtain a high throughput will also be discussed.

Wireless devices support multiple functionalities using a multitude of standards. In the WiFi ecosystem, for example, we have seen the addition of the new Gb/s flavors of 802.11ac and 802.11ad (WiGig). Hence, new WiFi-supporting devices have to deliver superior performance in each of these standards, and should operate seamlessly from standard to standard. These requirements demand replacing traditional designs with innovative solutions at system, architecture and circuit levels. In this work, we tackle the physical level challenges with a configurable transmitter architecture. This work demonstrates the concept of a single-PHY transmitter baseband architecture for 11ac and 11ad standards, as a supporting example. The core of the proposed transmitter is a configurable mixed-signal digital-to-analog converter, which has been fabricated in 28 nm FDSOI technology. It embeds semi-digital filtering tailored for four WiFi modes (20, 40, 80 and 160 MHz bandwidths) and the 1.76 GHz bandwidth of the 60 GHz WiGig standard.

The CubeSat platform was initially created by Jordi Puig-Suari and Robert Twiggs as an educational tool to provide a full end-to-end mission design and space-flight experience to university students pursuing two to four year undergraduate and graduate level studies. Since the CubeSat’s invention and subsequent release of the CubeSat Design Specification it has been widely and successfully adopted by many educational and research institutions. These institutions have used CubeSats to space qualify new technology and provide new science measurements to many under-served research communities. Recently, CubeSat’s have gained more traction in commercial and other governmental sectors focused on operational use. This has driven a need for new capabilities that still adhere to CubeSat standards while keeping with the risk tolerant design nature that allows for innovative mission designs and technology demonstrations.

During the past ten years, the number and type advanced device technologies deployed in the space environment has increased dramatically. This trend has been magnified by the number of organizations designing, building, and deploying platforms to low-Earth orbit and other locations in the Solar System. This increase has been driven by many factors, including ever-improving radiation-hardened by design (RHBD) techniques as well as serendipitous radiation tolerance in advanced technology nodes. Furthermore, many current and future aerospace applications demand capabilities that can only be realized with device feature sizes at and below 40 nm, 3-dimensional integration, and other techniques considering the substantial constraints placed on size, weight, power, and cost. In addition to application-specific integrated circuits (ASICs) and more traditional radiation-hardened components, many aerospace organizations are also expanding their use of automotive-grade devices and other enhanced commercial-off-the-shelf (COTS) offerings in order to expand the design space while trying to constrain risk. While advanced semiconductor device technologies enable many critical applications, their utilization in the space environment requires special considerations.

Silicon metal-semiconductor field effect transistors (MESFETs) can be fabricated using commercial SOI CMOS technologies without changing the CMOS process flow. With no fragile gate oxide, the silicon MESFETs can be designed to have breakdown voltages in excess of 30V, greatly exceeding that of the baseline MOSFETs. MESFETs with gate lengths as short as 150nm have current drives of >1A with fT and fmax greater than 30 and 45GHz respectively. They are also radiation tolerant (TID>300krad) and capable of operating over a wide temperature range (-180° to +150°C) making them ideally suited for space applications in extreme environments. This presentation will focus on the DC and RF characteristics of MESFETs fabricated using a 45nm SOI CMOS technology. Data from total ionizing dose measurements will be presented along with a wide temperature range TOM3 Spice model. The ability to integrate enhanced voltage MESFETs on the same die as ULSI CMOS has the potential for significant savings in size, weight and cost for space electronics, with the added benefit of increased reliability. As demonstrators for the integrated 45nm SOI CMOS-MESFET technology we will describe results from an unconditionally stable low-dropout regulator for point-of-load DC power management, and an RF power amplifier with Pout > 1W.

The challenge of developing state-of-the-art CMOS microelectronics for space applications is mitigating the effects of the space radiation environment. Radiation-hardened-by-Design (RHBD) has been demonstrated as an effective approach to leverage advances in commercial integrated circuit fabrication to provide improved performance, power, availability, and reliability for space applications. RHBD has demonstrated continued success in using design methods to achieve high levels of radiation hardness. New approaches to hardening mixed-signal designs that take advantage of the properties of modern CMOS processes have been developed. Examples of applications, development considerations, and analysis techniques will be covered in this session.

In this discussion we will first introduce the exciting Earth science, planetary science and astrophysics investigations that are performed by JPL and NASA at millimeter-wave and terahertz frequencies, describing several recent results by instruments operating in this wavelength regime. Then we will then discuss the important role CMOS system-on-chip (SoC) technology now plays in these instruments, and the fundamental challenges (noise, multiplicative-effects, radiation effects) that CMOS based instruments face in delivering the level of fidelity required for NASA’s science investigations. The talk will discuss two examples of CMOS SoC based instruments from recent NASA programs including a 600 GHz side-band separated spectrometer for investigation of Europa, Titan, Enceladus, and a 100 GHz in-situ spectrometer system for investigation of comet and asteroid volatiles targeting the planned NASA Comet Surface Sample Return (CSSR) mission.

RHBD is a not just a library or physical solution, in order to address mission goals (radiation, extreme environments compared to commercial), a system-level approach needs to be considered in the SoC. Understanding the use environment, both from a reliability and radiation response, drives the optimized design approach at both the macro/standard cell, and logic/behavioral levels. Physical modifications are required for some missions and we have migrated commercial IP to work in space applications in many technology nodes (multiple approaches to achieve a range of hardness levels). Integration of functions covering multiple frequencies from RF to KHz, also presents design challenges while maintaining space application goals. A discussion of these trades and approaches will be discussed.

This presentation discusses the benefits associated with the design of modular, scalable mm-wave PolyStrata™ based circuits for commercial and aerospace applications. The additive PolyStrata™ process technology was developed under the 3D-MERFS DARPA-funded research program to improve the performance and reduce the packaging cost of mm-wave systems. The technology characteristics include low loss, high component density, 3D-stacking and high isolation. 15 years later, this technology has become a key enabler for next generation Millimeter Wave (MMW) radar and communications systems. Presented herein are examples of a variety of modules, ranging from ultra-compact, low loss mm-wave filters qualified for space applications to complete PolyStrata phased array developed for NASA next generation instruments for remote sensing. Also discussed is the incorporation of PolyStrata™ passives components in T/R modules that improve the efficiency and increase power output.

Ohm’s Law forces the efficiency of linear amplifiers to remain below 50%, often well below 50%, leaving any practical amplifier achieving higher efficiency to not use linear circuitry.  Many such proposals have been put forth over the past century.  These are briefly reviewed and collected into 3 major groups.  One major impediment to efficient RF power amplifiers is the signal modulation types that have been selected by the Standards working groups.  Why this is so, and what physics allows in getting around this problem, is discussed.

High PAPR signals such as LTE-A cellular communications present a potential poor trade-off between efficiency and linearity. Power supply modulation is one of several techniques that can avoid this trade-off, and achieve high efficiency and with adequate linearity. This workshop presentation will focus primarily on the polar modulation method for both base-stations and mobile terminals (e.g. smart phones). Si circuits for supply modulators will be discussed, and how they can be applied to various classes of microwave power amplifiers (e.g. Class AB, F, etc). Both analog and digital envelope power modulators will be compared. Linearization methods will be presented tailored to this technique. The challenge of modulation bandwidth limitation will be addressed with several solutions. Finally, multi-carrier systems with potential digital beam steering applications using envelope-of-the-envelope techniques will be shown.

Digital Power Amplifiers (DPAs) such as the switched capacitor PA (SCPA) have provided a significant level of flexibility in SoC design. SCPAs provide a flexible, linear output, while also providing a digital interface that can interact directly with the DSP. In this talk, we will detail several SCPA architectures and provide example designs and some pros and cons for their use in varying wireless transmission applications.

An outphasing digital transmitter including open-loop delay-based phase modulator and class-D switching PA is presented. The PA uses a transformer power combining configuration with reduced losses at back-off power; class-D PA operation takes advantage of switching speed offered by CMOS technology scaling to achieve good linearity performance. System level topics, including digital front end line-up and phase modulation, will be discussed. Measurement from a 32nm test chip of the TX operating in 2.4GHz band with WiFi 20/40MHz signal will be shown as well as system simulation for 5-6GHz band with 80/160MHz signals.

Doherty power amplifier (PA) architecture offers a unique back-off PA efficiency enhancement behavior, which is highly desirable for amplifying broadband modulation signals with high peak-to-average power ratios (PAPR). Recently, there is an increasing interest to employ Doherty PA architectures for a wide variety of mobile communication and wireless connectivity applications. In this talk, we will first present an overview of benefits and design challenges of Doherty PAs. We will then present several CMOS mixed-signal Doherty PA examples that leverage both digital PA operations and analog PA techniques. These mixed-signal Doherty PAs enable precise controls of the Doherty main/auxiliary amplifiers to achieve optimum Doherty "active load-modulation" and optimum back-off efficiency enhancement. Moreover, the reconfigurability of these mixed-signal Doherty PAs enables PA linearity enhancement and robust Doherty performance under antenna load variations. These mixed-signal Doherty PAs can also be combined with other PA techniques to achieve hybrid PA architectures for further PA efficiency and linearity enhancement. In addition, we will also present the use of Doherty architecture in mm-wave linear PA designs for multi-band 5G applications. Supporting multiple 5G bands, such mm-wave linear Doherty PA enhances average PA efficiency, relaxes thermal management in 5G massive MIMO systems, and amplifies high-order QAM modulations without any digital-predistortion (DPD).

High bandwidth signals such as 802.11ax for wifi and future 5G networks demand a very high linearity and memory less PA. This very high spec is for the first time not only challenging CMOS PAs but also all GaAs and SiGe PAs as well. In this discussion subject of memory effects and its mechanism that leads itself to serious loss of efficiency and power delivery in high power PAs is explored.

The performance of efficiency optimized amplifiers (Class AB, Class J, Doherty or Envelope Tracking) can be enhanced significantly by digital signal processing algorithms, especially in linearity (EVM) and spectral regrowth (ACLR). At one level, performing crest factor reduction (CFR) assists the entire Tx chain, by not pushing the blocks to the signal peaks (to unreasonable levels) only to saturate the power amplifier (PA) further down the chain. There various digital predistortion (DPD) techniques that are used to compensate for PA distortion: memoryless compression, compression with memory, coordination of multiple drive signals, just to list a few.  These algorithms, strategies and approaches will be reviewed in this workshop talk.

This talk will give an overview of current ULP wireless solutions in the healthcare domain as well as present an outlook to potential future trends. First different wireless application areas in the healthcare domain, both for in-body and around-the-body use cases, will be presented. Second, a summary of existing ultra low power wireless communication technologies for those as well as an outlook to emerging wireless technologies in the field will be given.

The body channel communication (BCC) which uses the human body as a communication channel hase been getting more and more attention due to its good transceiver performance compared with traditional RF communiation including narrow band (NB) or ultra wide band (UWB). Thanks to the good performance of the BCC, the BCC was included in the IEEE 802.15.6 WBAN standard (which is called human body communication, HBC). In this talk, I will introduce you the BCC including the communication principle, channel analysis, optimized transceiver design and implementation results. Also, we will have a chance to discuss the future application of the BCC.

Radar technologies have been recently investigated in healthcare of which the general public will benefit in terms of diagnostics, treatment, and detection of emergency situation. They represent the new emerging solution to promote the health both in home and clinical environments, and are also predicted to proliferate in the next years. This is in line with the development of the Internet of the Things. Although such remote sensing has the limitation that only biomedical parameters that are based on mechanical movements and/or distance can be monitored, radar operation allows characterizing already various biomedical parameters of strong interest in several applications with the great advantage on being non-invasive. This presentation discusses system requirements, design challenges, practical limitations, and solutions of recent advances in health monitoring using radar technologies, presenting also some experimental results.

Implantable medical devices such as devices for Cardiac Rhythm Management (Pacemakers and ICD’s) and Neuromodulation treat a wide range of medical conditions and have benefitted tremendously from the advancement of wireless communications. The recent introduction of Bluetooth Low Energy (BLE) communication capability in these devices offers further value from both the patient and physician’s perspective. This presentation will touch on some of the benefits that BLE provides while detailing aspects of the Bluetooth Low Energy standard. Some of the technical challenges of implementing a BLE transceiver in an implantable medical device will be highlighted, such as power constraints and performance requirements. Finally, this presentation will conclude with a discussion on future opportunities made possible with BLE communication.

Over the last two decades, computers have evolved from the laptop to the smartphone to today’s cm-scale IoT devices. At each step, the volume has reduced by 2-3 orders of magnitude, and with it so has the size of the battery. However, the functionality has remained constant or even increased, a trend that shows little sign of slowing. The next logical step in computing are millimeter-scale devices with similar features found in today’s smartphones. With thin-film batteries, CMOS scaling, low-power sensors, and advanced packaging, these mm-scale devices have now become a reality. This presentation focuses on recent advances in wireless communication circuits, antennas, and protocols for mm-scale sensor nodes, for implantable and on-body applications. Challenges resulting from miniature antennas, thin-film batteries, and operation from a <100nW harvested power budget will be discussed, along with solutions that have been fabricated and demonstrated in complete, autonomous mm-scale sensor nodes.

An energy-efficient radio SoC with RFFE (RF front-end), DBB (digital baseband) and MCU (microcontroller) for medical/ healthcare applications in the 315/400 MHz bands is presented. The SoC is compliant with the IEEE 802.15.6 standard in the 402-405MHz MICS (Medical Implant Communication Service) band and the 420-450MHz medical telemetry band, and also supports a proprietary high data rate mode (3.6Mb/s) to support applications such as EEG (electrocardiogram). The ADPLL (all-digital phase-locked loop)-based two-point modulation transmitter is adopted to support wideband modulation beyond the PLL bandwidth. The total power consumption of RX and TX are 3.5mW (3.6Mb/s, -77dBm sensitivity) and 3.0mW (3.6Mb/s, -16dBm PA output) respectively enabling energy efficiency of less than 1nJ/bit.

Bioelectronic modulation of neural activity has the potential to provide therapeutic control over diverse organ functions addressing unmet clinical needs.  Towards this goal, significant progress has been made in the development of miniaturized electronics, and high resolution and mechanically flexible neural interfaces for both research and clinical systems.  Their long-term access to neural structures, however, remains constrained by technological challenges in powering the device.  In this talk, I will describe two new methods for electromagnetic energy transfer that exploit near-field interactions with biological tissue to wirelessly power tiny devices anywhere in the body.  I will discuss engineering and experimental challenges to realizing such interfaces, including a pacemaker that is smaller than a grain of rice, a conformal vagus nerve stimulator, and a fully internalized neuromodulation platform.  These devices can act as bioelectronic medicines, capable of precisely modulating local activity, that may be more effective treatments than drugs which act globally throughout the body.

Next generation communication systems (5G and beyond) seek to bridge the gap between the projected demand and supply of mobile data traffic through a combination of new system techniques and access to new spectrum below 6 GHz and especially in several millimeter-wave bands from 15 GHz to 86 GHz. In these systems, the design of frequency synthesizers that can access several such bands with low phase noise, spur levels and frequency granularity remains a critical block. Furthermore, “Massive MIMO” – which consists of a large number of antennas at the access point – is a promising technology to meet the high data rate and quality of service requirements of 5G wireless systems. Achieving stringent phase-noise specifications and scalable LO distribution to maintain phase coherence across the different units in the MIMO array is a critical challenge. This workshop will present the latest trends in the design of such synthesizers.

Modern wireless communication systems require agile support for an increasing number carrier frequencies. The objective of this work is to build a single PLL which is sufficiently flexible such that it can be used in a field programmable radio, or a radio system capable of being reconfigured in the field to any of a diverse set of radio communication standards. An obstacle to producing such a reconfigurable radio is that the noise, frequency, and bandwidth requirements for different standards vary substantially, in turn necessitating multiple PLLs, each optimized for a particular application. To meet the goals of our research effort, it is necessary to produce a PLL which is reconfigurable in multiple dimensions yet maintains sufficiently strong performance (in terms of metrics such as phase noise and locking time) such that the resulting synthesizer is still competitive with the best-performing single-application CMOS PLLs, and can therefore effectively replace a set of application specific PLLs. This presentation will provide an overview of recently demonstrated VCO and PLL designs/architectures which have made the above-described vision of a highly flexible monolithic PLL possible.

The phase accuracy of CMOS circuits for clock generation and distribution is fundamentally limited by component mismatch and noise. Both phase mismatch and phase noise, and hence timing jitter, can be improved by admittance scaling at the cost of power consumption. To benchmark and optimize circuits in a power efficient way, it is useful to define a Figure of Merit (FoM) that normalizes for this admittance level scaling effect. Based on this Jitter-Power FoM several circuit techniques for accurate clock generation will be discussed, i.e.: 1) Multi-phase clock generation, comparing a delay locked loop versus a shift register or divider 2) The choice of the logic family, comparing dynamic transmission gate logic with current mode logic 3) Digital to Time Conversion (DTC) and its use in frequency synthesis

Modern large scale SOCs combine high performance radio IP along with large ASIC content. Meanwhile the requirement for high performance PLLs put more constraints on the design with limited power supply choices, close proximity to noisy ASIC blocks and large variation of junction temperatures. Multiple standards requires multiple PLLs to coexist on the same SOC. Balancing the often conflicting needs among different IP blocks requires the frequency synthesizer to be designed and optimized at architecture level. This talk covers design considerations in the frequency synthesis of the modern CMOS SOC design including tradeoff in frequency coverage, tuning range, spur and coupling between different PLLs.

Emerging wireless networks at RF and mm-wave are addressing the demand for higher network capacity by focusing on multiple-element arrays for interference mitigation and for meeting challenging link budgets. Maintaining coherence between elements is critical for such phased-arrays/MIMO systems, particularly as arrays evolve towards large number of elements such as massive MIMO base-stations. In addition, wide range of operating frequencies require wide LO tuning range. Hence, scalable, power-efficient, frequency generation and distribution approaches that support low phase noise and wide tuning range for large-element arrays are of interest at RF and mm-wave. In this talk, we will present a broad overview of approaches for LO generation and phase coherence in large-scale arrays. Techniques for achieving wide tuning range will be presented along with trade-offs associated with traditional clock distribution schemes. A scalable clock distribution approach that ensures low phase noise and jitter with increasing number of elements will also be presented in the context of an implementation targeted at emerging mmwave 5G networks.

Advanced design techniques for CMOS mm-Wave phased-array frequency synthesis will be presented. Firstly, in a 4-path LO generation system for 60-GHz phased-array receivers, a frequency tripler with a proposed locking-range enhancement technique is proposed to relax both the frequency and the phase shift of injection-locked-oscillator-based phase shifters for high linearity and low power. Fabricated in a 65-nm CMOS process, the 4-path LO generation system measures linear phase range > ±90°, amplitude variation < ±0.4 dB, phase resolution of 22.5° while drawing 85 mA at 1V. With a successive-approximation algorithm to perform automatic phase detection and tuning, the maximum phase errors is reduced from 22.0° to 1.5°. Secondly, a 0.6-V 14.1-mW 100-GHz 65-nm transformer-based phase-locked loop (PLL) with embedded phase shifters is demonstrated with phase resolution of 3.9°, amplitude variation < ±0.1 dB, and phase noise of -88 dBc/Hz at 1-MHz offset from 100 GHz.

In All Digital PLLs for Frequency Synthesis efforts are made in designing the TDC/BPD and DCO with low quantization noise and high performance capabilities. Minimum effort has been put in to the design of the loop filter. Mainly, the Filter remains a digital representation of the Analog implementation used in the well known Analog PLLs. This session will analyze from a theoretical point of view the required characteristics of a digital Loop filter to maximize ADPLL performances when a Single Bit phase detector is used.

Next generation communication systems (5G and beyond) seek to bridge the gap between the projected demand and supply of mobile data traffic through a combination of new system techniques and access to new spectrum below 6 GHz and especially in several millimeter-wave bands from 15 GHz to 86 GHz. In these systems, the design of frequency synthesizers that can access several such bands with low phase noise, spur levels and frequency granularity remains a critical block. Furthermore, “Massive MIMO” – which consists of a large number of antennas at the access point – is a promising technology to meet the high data rate and quality of service requirements of 5G wireless systems. Achieving stringent phase-noise specifications and scalable LO distribution to maintain phase coherence across the different units in the MIMO array is a critical challenge. This workshop will present the latest trends in the design of such synthesizers.

mm Wave frequencies are a new frontier for 5G but have lack of clarity until WRC2019. The use cases should define where mm Wave plays a role. A prominent example is the 2018 Olympics in Korea intend to use 28GHz. Nevertheless there are challenges ahead in regulation, standardization and implementation which require common efforts from the whole value chain to make it happen.

In the last two decades data-rates in wireless communication systems have been increasing exponentially. This trend is continuing with the fifth generation of wireless systems (5G) that will require peak rates in excess of Gb/s for many users, several hundred thousands of simultaneous connections for massive sensor deployments, and substantially improved spectral efficiency. This workshop is focused on current state-of-the-art of 5G band and future directions of the key circuit techniques and system architectures for base station or between the Handset, or other portable devices, and the cell, or mini-cell, micro-cell, pico-cell base stations. All aspects covering normalization Systems, Architecture, and low power design solutions for beam orientation will be discussed.

In this presentation we focus on trends of power amplifiers and PLLs being two of the most discussed RF components within 3GPP towards standardization of 5G at mmW frequencies. The decreasing PA output power and power-added efficiency for increasing carrier frequencies challenge the choice of waveforms and spectral emission requirements. PLL performance degrades with increasing carrier frequencies and affects attainable error vector magnitude and the choice of the 5G numerology. Furthermore, the use of array antenna systems puts additional constraints on RF components, e.g. in terms of accuracy and tracking over the array. We discuss the present state of 3GPP 5G specifications (Release 15) and its implications on requirements for power amplifiers and PLLs.

The rapid growth of wireless data drives new design challenges for RF chipset and handheld/mobile device manufacturers along with carriers. The massive data traffic to be supported by wireless networks requires the development of cost effective high speed and low power wireless link (both from end user and network side). To address this challenge, millimeter wave technologies (WiGig standard at 60 GHz, backhauling in E band …) have emerged as promising solutions in order to offer multi gigabit per second data rate at low power. Moreover, the possibility to integrate those wireless systems using silicon based technologies (either CMOS or BiCMOS) enables to offer cost effective solutions required by consumer markets. But millimeter wave technologies do not only require cost effective RF ICs achieved in advanced silicon technologies, high performances and low cost packaging and antenna technologies are also key issues. This talk will illustrate how industrial organic packaging technology and standard FR4 based HDI any layer PCB technology (used for smartphone board) can help to develop innovative and cost effective mmw packages and antennas (focusing here on 60 GHz applications).

In the last two decades data-rates in wireless communication systems have been increasing exponentially. This trend is continuing with the fifth generation of wireless systems (5G) that will require peak rates in excess of Gb/s for many users, several hundred thousands of simultaneous connections for massive sensor deployments, and substantially improved spectral efficiency. This workshop is focused on current state-of-the-art of 5G band and future directions of the key circuit techniques and system architectures for base station or between the Handset, or other portable devices, and the cell, or mini-cell, micro-cell, pico-cell base stations. All aspects covering normalization Systems, Architecture, and low power design solutions for beam orientation will be discussed.

In this presentation, possible 5G frequency bands will be discussed from the viewpoint of device technology, data rate, communication distance, array size, etc. Some millimeter-wave CMOS transceiver will be also introduced, achieving tens of Gbps data rate.

This talk will highlight both phased-array front-ends and tunable receivers for use in mm-wave 5G systems. One key requirement for mm-wave arrays is achieving improved power efficiency. For basestations, this can reduce cooling requirements whereas for user equipment, this can be translated into reduced array size. With this goal in mind, harmonic-tuned SiGe power amplifiers at 28-GHz will first be presented which achieve 35% peak PAE and up to 13% PAE at 6 dB back off. To further improve back-off efficiency, a dual-vector Doherty beamformer is introduced and demonstrated at 60 GHz in SiGe BiCMOS. This beamformer achieves 7% PAE at 6-dB back-off and allows for a reconfigurable linearity/efficiency trade-off. Another goal for mm-wave systems is achieving compact, low-power, and widely tunable performance. Towards these ends, a mixer-first receiver will be presented which can be tuned across 20 to 30 GHz and achieves 7-8 dB noise figure at 41 mW power consumption.

The fastest growing wireless communications market has seen dramatic changes in both the requirements on, and the capabilities of the radio to support wireless connections. Tunable RF technologies are enabling new frontiers for reconfigurable RF front ends. Miniaturized multifunctional frequency-agile devices are demanded to support multiple communication standards in current and next generation military and commercial applications. While active components in RF front-ends are experiencing higher levels of co-integration, there is a technological barrier for further integration to achieve miniaturized communication systems due to large amount of discrete RF passives on board. This talk will present the development of frequency agile and fully electrically tunable miniaturized RF devices with the integration of ferromagnetic and ferroelectric thin films and related techniques enabling multifunctional and adaptive radios. First, the need for adaptive and reconfigurable frequency control components in next generation wireless devices will be described. Previous tuning techniques (e.g., RF MEMS) and their performance will be briefly reviewed, followed by application of these components in adaptive wireless systems. Design of novel slow wave structures and their RF applications will be discussed for further miniaturization of RF passives. And finally, integration of slow-wave structures with nano-scale ferromagnetic and ferroelectric (PZT) thin film patterns will be demonstrated and investigated for their efficacy in developing fully electrically tunable RF components (e.g., filter, antenna). Strategies to improve RF characteristics (e.g., Ferromagnetic Resonance Frequency and tuning range) of the integrated thin films will be fully addressed and discussed.

With ever-increasing development of the modern communications (i.e., 5G wireless, internet-of-everything, and so on), compact and practical transceivers to meet the requirements of such system remain as great challenges. Thus, integrated passive circuits with high performance as key elements are dramatically demanded and highly developed based on novel miniaturized structures and specific technologies, which can be utilized for the implementation of the RF, microwave, millimeter-wave, and THz communication systems. In this talk, novel integrated passive circuits for RF, microwave, millimeter-wave, and sub-THz applications are introduced for modern communication with enhanced performance. First, the on-chip stepped-impedance inductor with an adjusted high quality-factor is analyzed and employed for RF VCO design with low phase-noise performance. Secondly, based on such type inductor, a stacked stepped-impedance transformer is implemented for the wideband matching network with high passive efficiency for a microwave power amplifier design with a wideband operation (i.e., 3.5-9.5 GHz). Thirdly, a novel on-chip 3D embedded capacitor is introduced for mm-wave application, i.e., 60 GHz DCO. Finally, a new slow-wave sub-THz resonant cell is introduced for the dual-resonance (i.e., 237 and 380 GHz) allocation for the dual-band sub-THz application. All the passive circuits mentioned above are fabricated using the silicon-based technology (i.e., CMOS and SiGe). Good agreements between the measurements and simulations are achieved, which verify the feasibility of the proposed circuits for the practical applications.

Future big-data oriented commuting requires energy efficient I/O links for data migration. Silicon photonic interconnect has great performance for each individual optical component under different process technology but has limited performance after integration. This talk will address this challenge by exploring electrical interconnect solution at Terahertz with source, transmission and detector all realized in CMOS. However, the big problem here is the poor output power of source and huge crosstalk of transmission at Terahertz in CMOS. We show that with the use of meta-device (meta-material, meta-surface, spoof surface-plasmon-polariton), one can effectively manipulate the EM-field such that in-phase power combing can be utilized to combine output power (3-5dBm) of coupled oscillators; and surface-plasmonic wave can be utilized for low-crosstalk (-21dB) on-chip transmission. Demonstrated chips with measurements at 140GHz and 280GHz by standard 65nm CMOS process. What is more, the utilization of meta-device for imaging and sensing at THz in CMOS will also be briefly reported.

This talk will explore the major challenges for radio-frequency micro-electromechanical systems (RF MEMS) and ways to overcome them. MEMS is ubiquitous. For example, a smartphone can have many MEMS components including microphones/speakers, camera focus/vibration controls, micro-projectors, silicon clocks, position/motion sensors, pressure/humidity/temperature sensors, etc. In comparison, RF MEMS have not been widely used despite their promising performance. In fact, like many new technologies, RF MEMS followed the Gartner hype cycle to reach the peak of inflated expectations around 1995, only to crash to the trough of disillusionment around 2010. However, this also implies that they are now on the slope of enlightenment and may soon reach the plateau of productivity if the major challenges can be overcome in a timely manner. The initial hype of RF MEMS originates from their inherent advantages in terms of lower loss, higher isolation, wider bandwidth, better linearity, and lower power consumption compared to silicon CMOS transistors. However, the hype quickly turned into disillusionment mainly due to the reliability issue. Over the following decade, the reliability issue was mostly overcome by careful choice of design, material and bias conditions, more for capacitive switches than ohmic switches. The main challenge then turned into a yield issue, which was largely overcome by fabrication through standard CMOS foundries instead of dedicated MEMS foundries. Now, the main issue of RF MEMS is insufficient performance advantage to overcome the entry barrier faced by any new technology in competition with the entrenched CMOS technology, such as in existing 3G/4G smartphone sockets. In this case, the savior appears to be the necessity for 5G wireless systems to expand above 6 GHz, which can only enlarge the inherent advantage of RF MEMS, especially capacitive MEMS. However, RF MEMS need to quickly increase their integration level with CMOS circuits. Otherwise, RF MEMS will risk inheriting the curse previously placed on compound semiconductors as “the technology of the future and it will always be.”

Due to its low insertion loss, high isolation, high linearity, wide bandwidth, and near-zero dc power consumption, radio frequency micro-electromechanical (RF-MEMS) switches have been an emerging technology that can be used in automated test equipment (ATE), wide-band instrumentation, switching matrice, digital attenuators, satellite switching networks and defense systems to achieve superior system performance. Compared with capacitive RF-MEMS switches, metal contact type switches have large bandwidth from dc to RF frequency and are favored in many wide-band applications. However, the reliability issues associated with RF-MEMS contact switches have been a barrier for wider adoption of the technology. In particular, existing RF-MEMS switches exhibit poor hot-switching life-time due to dielectric breakdown and field emission. In this talk, we present several design methodologies for drastically improving the hot-switching reliability of contact-type radio frequency micro-electromechanical (RF-MEMS) switches. In the proposed design, sacrificial contacts are used together with low-resistance contacts to significantly reduce the electric field across the latter. The lower field strength drastically reduces the contact degradation associated with field induced material transfer. Theoretical and numerical modeling show that the proposed protection scheme introduces minimal impact on the switch’s RF performance. To realize the protection scheme, we introduce a novel mechanical design that allows the correct protection actuation sequence to be realized using a single actuator and bias electrode. As a demonstration, several 0–40 GHz RF-MEMS switches are fabricated using a robust copper sacrificial layer technique. Compared with unprotected switches, the protected switch design exhibits over 100 times improvement in hot-switching lifetime. In particular, we demonstrate 100–150 million cycle lifetime at 1W hot-switching and 50 million cycles at 2W hot-switching before catastrophic failure, measured in open-air lab environment. Further optimization of the structural design and contact materials is likely to further increase the hot-switching lifetime.

Integrated Passive Device (IPD) technology is a good platform for realizing compact high-Q passive components such as baluns, filters, couplers and transformers. In contrast to other multilayer technologies, that suffer from issues related to control of dimensions, the IPD technology allows the realization of relatively low cost passives with well controlled dimensions. The presentation shows examples of IPD passive devices fabricated in a commercial IPD foundry as well through the use of fabrication process developed in a university clean room. Examples of integrating IPD devices with BST technology to develop tunable IPD devices are also presented.

Recent advances in wireless communication systems call for RF front-ends with multi-standard and multi-functional operability which in turn create the need for RF filters with small physical size, low-loss and flexible transfer function. Acoustic-wave resonators have been the key technology of these systems due to their unique advantages of highly-miniaturized volume and high quality factor (Q). However, their operation is limited by narrow fractional bandwidth (FBW, << the electromechanical coefficient kt2), static response and limited type of realizable transfer function (mostly bandpass). In order to address the aforementioned drawbacks, a new class of acoustic-wave-lumped-element (AWLR)-based RF filters has been developed and will be presented at this workshop. They are based on a hybrid implementation scheme in which acoustic-wave resonators are effectively combined with electromagnetic-wave elements. They allow the realization of acoustic-wave-resonator-based filters with: i) enhanced FBW (>kt2), ii) small physical size, iii) high-Q (1,000-10,000) iv) arbitrary transfer function synthesis (i.e, bandpass, bandstop, absorptive) and iv) tunable response. Within the course of this workshop, various AWLR-based filtering topologies will be presented in terms of RF design synthesis (coupling matrix approach) and experimental performance with a major focus on bandpass- and bandstop-type responses. Moreover, the RF design of multi-band bandpass filters with various levels of reconfigurability will be reported. Absorptive-bandstop filters with highly selective and tunable equi-ripple stopbands will also be presented.

This talk will discuss about the great potential of RF and microwaves frequency micro-systems to achieve dielectric spectroscopy measurement on individual micro-particles and their promising use for biological cell analysis and discrimination. We will review the latest innovations and results achievements based on passive microwave microfluidic sensors that allows nowadays possible accurate dielectric properties characterization of biological sample up to single cells. Achieved electromagnetic signatures measurements may address some promising biological issues especially in cancer research. Hence, we will show that such measurements, without requiring any conventional cell labelling, could be used to discriminate cells and be helpful to early inform on potential cell aggressiveness degree for example.

Microwave resonators have proved their capability as sensing devices in a wide range of applications such as lab on chip, environmental sustainability, and industrial applications, not only for analyses in the solid and liquid phase, but also recently in a gaseous environment. Planar microstrip resonators made from split ring resonators (SRRs) have shown relatively sharper resonances and concomitantly increased sensitivity due to their higher coupling to the surrounding signal lines. These sensors are amenable to miniaturization, automation, mass production, and wireless interconnection due to CMOS compatibility, low cost and a facile fabrication process. Such microwave resonators also offer noninvasive sensing through contact-less probing, which adds to their flexibility of usage and maneuverability for in situ characterization. Since SRR based passive sensors have a low response time and can almost instantly translate changes in the environment of study into measurement quantities, they can be used as real-time sensors. Their unique capability of detecting small complex permittivity in non-contact fashion distinguishes them from other sensor technologies. In this workshop design and analysis of such sensors for gas and chemical sensing, biomedical applications and nanomaterial characterization will be discussed.

(Note: This is not the official abstract, but the workshop coordinators description of what the talk will cover - Nathan) This talk will discuss the different wireless standards, both present and future, that will be major factors in the IoT.

This talk presents holistic system analysis, design and optimization approaches to realize ultra-low power (ULP) communication for energy-optimized Internet-of-Things (IoT). A truly energy-optimal IoT wireless connectivity can only be obtained by a cross-layer optimization that requires a full characterization of the complete end-to-end system. Addressing this critical technical challenge in emerging ULP IoT applications, an interdisciplinary research that spans wireless communication, signal processing, and VLSI circuits will be discussed in this talk. Bridging the noticeable gap between communications/signal processing theories and VLSI circuit implementations, this talk introduces holistic solutions obtained by a rigorous theoretical analysis that considers practical implementation issues such as non-idealities in circuits and energy constraints in signal processing. To demonstrate the proposed cross-layer system design approach, an energy-autonomous wireless communication for ultra-small (i.e., millimeter-scale) sensor nodes will be presented. In addition, software-defined radio architectures for multi-standard-compliant IoT communication, WiFi / Bluetooth back-channel communication schemes to interconnect heterogeneous networks, and an ULP 100m-range cm-scale accuracy indoor localization solutions for RFID tags will be discussed.

This talk presents the progress of LPWA (Low Power Wide Area) technologies, especially NB-IoT. Why NB-IoT is the best choice of operators will be presented. How to construct a commercial NB-IoT from the existing telecom equipment will be introduced. The choice of frequency band will be introduced. An end-to-end total solution of NB-IoT is presented and different application cases of NB-IoT will be introduced as well.

(Note: This is not the official abstract, but the workshop coordinators description of what the talk will cover - Nathan) A discussion on multi-standard low power radios. These radios are designed to have low power and are capable of communicating over multiple standards.

Ultra-low-power IoT radios need to balance a difficult trade-off between energy efficiency and spectral efficiency to simultaneously enable long battery life while also enabling reliable communication in bands with severe spectral congestion. However, nearly all prior-art ultra-low-power radios communicate with spectrally inefficient schemes like OOK or FSK, as these modulation schemes facilitate low-complexity, energy-efficient designs. This presentation will explore the design of new transmitter and receiver architectures that enable high-complexity modulations at high efficiency. Importantly, designs will be restricted to CMOS to ensure low-cost radio integration into next-generation IoT devices.

The exponential growth in the semiconductor industry has enabled computers to pervade our everyday lives, and as we move forward many of these computers will have form factors much smaller than a typical laptop or smartphone. Sensor nodes will soon be deployed ubiquitously, capable of capturing information of their surrounding environment. The next step is to connect all these different nodes together into an entire interconnected system. This “Internet of Things” (IoT) vision has incredible potential to change our lives commercially, societally, and personally. The backbone of IoT is the wireless sensor node, many of which will operate under very rigorous energy constraints with small batteries or no batteries at all. It has been shown that in sensor nodes, radio communication is one of the biggest bottlenecks to ultra-low power design. This research explores ways to reduce energy consumption in radios for wireless sensor networks, allowing them to run off harvested energy, while maintaining many qualities that will allow them to function in a real world, multi-user environment. Three different prototypes have been designed demonstrating these techniques. The first is a sensitivity-reduced nanowatt wake-up radio which allows a sensor node to actively listen for packets even when the rest of the node is asleep. CDMA codes and interference rejection reduce the potential for energy-costly false wake-ups. This radio consumes 116nW active power, which is less than the sleep power of a standard sensor node, while achieving a data rate of 12.5kbps and an energy/wake-up of 287.7pJ. The second prototype is a full transceiver for a body-worn EKG sensor node. Due to size and energy limitations, this transceiver is designed to have low instantaneous power and is able to receive 802.15.6 Wireless Body Area Network compliant packets. It uses asymmetric communication including a wake-up receiver based on the previous design, a 4µW UWB transmitter and a communication receiver that consumes 292µW. The communication receiver has 10 physical channels to avoid interference and demodulates coherent packets which is uncommon for low power radios, but dictated by the 802.15.6 standard. The third prototype is a long range transceiver capable of >1km communication range in the 433MHz band and able to interface with an existing commercial radio. A digitally assisted baseband demodulator was designed which enables the ability to perform bit-level as well as packet-level duty cycling which increases the radio's energy efficiency. At 2.5mA transmit and 378µW receive power, the transceiver is more than 20X more efficient than its commercial counterpart.

(Note: This is not the official abstract, but the workshop coordinators description of what the talk will cover - Nathan) A discussion on battery-free computing and communication. Topics include backscattering and creating devices that can leverage existing wireless spectrum to communicate for ultra-low power.

Automotive radar and thus ADAS based on radar has been in development over decades – E.g., Mercedes-Benz went into series production with 77 GHz radars for premium cars in 1998, until mid 2015 Infineon had sold 10 million 77 GHz chip-sets worldwide, Hella has delivered 20 million 24 GHz BSD (Blind Spot Detection) units until July 2016, or Mercedes-Benz installed more than 3,5 Mill. radar units in 2015. New cars out of today’s production are equipped with a lot of different ADAS units - based on radar, lidar or cameras. Specifically the semiconductor industry there are many challenges that are actively worked on. A system level hot debate is ongoing on the subject of central vs. de-centralized processing between the silicon supplier and integrators. Some OEMs, like Audi or GM, are preferring standardized central processing (zFAS – zentrales Fahrer Assistenz System) while companies like Infineon, NXP, TI or NI, as well as ADI are fostering the de-centralized approach. A topic of similar impact on the IC design is the choice of technology with a strong push to full CMOS solutions, instead of staying with SiGe for the RF. Another relevant subject is the entire area of sensor measurement – in production as well as in the after sales world. Above mentioned topics as well as other system level topics from the car manufacturer's perspective will be addressed is this talk.

In the last few years the automotive safety market gets more and more important especially radar based systems like adaptive cruise control, blind spot detection or emergency brake systems. To be able to handle the higher demand of automotive radar systems and to reduce the system costs at the same time, it is necessary to increase the level of integration, especially on the high frequency (HF) front end side. Traditionally the radar market was dominated by HF bipolar semiconductor based transceivers. For a higher level of integration and to fulfill the high demands of HF performance different technologies like bipolar complementary Metal-oxide-semiconductor (BiCMOS) are necessary. This shift in technology favored a more “digital” centric transceiver partitioning, although the traditional analog HF-architectures (e.g. HF VCO, power amplifier, HF mixers) remained in bipolar technology. It also offers the possibilities of new transmit modulation concepts, and especially in digital front end (DFE) enhanced receiver architectures. Another important aspect for higher integration is the growing demand on monitoring functionality with high coverage to fulfill the requirements for an ISO26262 compliant design. This and more conceptional aspects including an outlook will be part of this presentation during the workshop.

The talk will present the latest development in automotive radar phased-arrays at UCSD with emphasis on 77-81 GHz systems. We have built 32 and 64-element FMCW radar phased-arrays capable of high resolution imaging, and which allow for ADAS functions. The talk will present measured patterns and some recent radar results.

Vehicular radar is perhaps the most compelling application of silicon-based mmWave circuits and systems. While multiple-antenna systems, such as phased arrays, have been explored for mmWave vehicular radar and enable operation under weak-SNR conditions, the potential of communications-inspired MIMO techniques as applied to radar (or the so-called MIMO radar concept) has yet to be significantly explored. This presentation will initially discuss MIMO radar principles at the system-level, including space-time array processing, multi-beam MIMO radar, waveform trade-offs etc., and will then move on to CMOS implementations in the 22-29GHz frequency range.

tbd

In spite of the very-high frequency of the application, the automotive industry is demanding state-of-the-art performances while ensuring low-cost and reliable solutions at the same time. To comply with this request silicon manufacturer are increasing the level of integration of MMICs embedding more and more functions in a single die. In this talk, we discuss advanced transceiver solutions in SiGe BiCMOS technology for both 24 and 77 GHz automotive radar applications. A general overview of architectures, circuits, and package solutions is described reporting recent products key performance. Finally, an outline of next challenges will be shortly discussed to provide the audience with the trend in this field of applications.

The increased interest in compact and low power automotive radar sensors toward fully autonomous vehicles pushes the research in 79GHz radar. While frequency and phase modulated radars in CMOS and SiGe have been successfully demonstrated, most of them do not integrate the mm-wave parts together with the necessary digital signal processing. This talk will discuss the integration of mm-Wave radar transceiver based on PMCW in advanced CMOS.

tbd

tbd

tbd

tbd

An All-digital PLL architecture targeting Wide-Area IoT Cat-M and NB-IOT standards enabling power transmitter co-integration exploits back-gate control in FD-SOI. Subsystem architecture through block designs achieve low power, small size and ease of SoC integration. A Digitally controlled 2.8-4 GHz oscillator prototype achieving a step size of 12kHz and 188 dB FoM, provides a basis for realizing this fully integrated 22nm subsystem.

This presentation will discuss the main features of FD-SOI CMOS technology and how to efficiently use its unique features for RF and mm-wave SoCs. We will overview the impact of the back-gate bias on the measured I-V, transconductance, fT and fMAX characteristics and compare the MAG of FDSOI MOSFETs with those of planar bulk CMOS and SiGe BiCMOS transistors through measurements up to 325 GHz. Finally, we will provide examples of LNA, mixer, switches, and PA circuit topologies and layouts that make efficient use of the back-gate bias to overcome the limitations associated with the low breakdown voltage of sub 28nm CMOS technologies.

RF data converters, currently in 16nm FinFET, for wireless and wired infrastructure applications. As well, this talk will cover low phase noise clock generation and distribution circuits

Designing highly sensitive RF/Analog circuits to operate in a hostile environment with extreme level of switching noise pose a huge challenge. This workshop presents a case study of a cascade PLL system for wide-band SerDes applications that support data rates from 1.25Gbps up to 26Gbps. The PLL system were implemented in the latest 10nm FINFET process technology. The various aspect of design issues were examined in this study, such as high flicker noise, switching noise interference, self-heating, low voltage supply, high mod-sense of FINFET varactor, etc.

For advanced node integration, small, inductor-free LNAs offer cost and SoC co-integration advantages. Exploiting the high self-gain and fT of FinFETs to minimize noise figure and current consumption while accommodating FinFET gate resistance presents a new design challenge. Noise cancelling topology benefits and trade-offs will be explored through a complete 14nm LPP-XL design example, through layout and extracted view results

FinFET process technologies offer lower chip area and lower supply voltages compared to planar CMOS devices. We will present circuit architectures, design challenges and measurements on RF building blocks (receivers, transmitters, synthesizers) designed in a 14nm FinFET technology for 4G (LTE) wireless cellular communication standards.

This talk will review what we actually mean when we say a measurement is uncertain. The talk will examine, in detail, some of the processes that can make measurements appear uncertain. The talk will also discuss the two main methodologies that are currently being used for evaluating the size of a measurement’s uncertainty. Both these approaches are currently recommended in international guidance documents for evaluating measurement uncertainty (i.e. as published by the International Organization for Standardization, ISO). However, these two approaches are incompatible with each other. The talk will show that the two approaches can produce very different values for the uncertainty of a measurement – in particular, for many measurements that are made at microwave frequencies.

Usually EM vendors try to say how accurate their software is. But accuracy is not quite what the skilled microwave designer wants to know. Rather, they are interested in what the error is. To get this information, just like the civil engineer who builds a model of a bridge only to stress it to the point that it breaks, we demonstrate how to break EM software. In doing so, we find out that all EM software always gives the wrong answer, without exception. In order to realize success with high probability, we need to establish firm upper bounds on that error and learn how to keep those upper bounds low enough that we have a good chance of success. Probability cannot be left to chance.

Today microwave techniques have an enormous impact on a huge number of fields from health care to telecommunication. This plethora of applications is achieved with more and more integration of microwave functionalities into general purpose systems thus the design and measurement techniques have been pushed and pushed toward the holy grail of first run IC success. However, the characterization and the measurements are intrinsically affected by an uncertainty, which is a function not only of the system but also of the DUT. For this reason a new Real-Time uncertainty approach able to quantify and present not only the nominal value, but also the associated uncertainty is becoming more and more fundamental for the high level of accuracy required for modern microwave system. The talk will present the recent advance in the field and in particular the methodology and the results achieved with a new real-time uncertainty computation engine.

An essential step after any calibration measurement is verification of the calibration accuracy. For the verification of linear S-parameter calibrations, several procedures and best practices have been proposed and developed in the past. Such procedures typically rely on well-known linear calibration standards (short, open, load, etc.). In large-signal Nonlinear Vector Network Analyzer (NVNA) measurements, however, linear calibration standards are insufficient and the development of a stable and accurate nonlinear verification standard has been a hot research topic for several years now. One of the main scientific challenges is to make the performance of the nonlinear verification device insensitive to the surrounding measurement instrument itself (for example, test port mismatches and DC bias variations). This talk will review state-of-the-art nonlinear verification standards and present the design of a new and promising nonlinear verification device. It will be shown how non-ideal instrument characteristics such as test port mismatches and DC bias variations influence the device performance. Load-pull measurement results and an improved Figure-Of-Merit (FOM) are furthermore presented and used to qualitatively compare device performance against state-of-the-art.

The presence of electrical noise is a common source of uncertainty, especially in the case of high bandwidth and low signal amplitude level. The development of high performance nanoelectronic integrated circuits and systems with increased functionality, high bandwidth and low signal power levels requires EMI-aware design. Such a design has to be based on accurate signal- and noise modeling to minimize the uncertainty in the operation of the circuits and systems. In this contribution we present advances in measurement and modeling of noisy electromagnetic fields using two-probe scanning and correlation analysis. Stochastic electromagnetic fields with Gaussian amplitude probability distribution can be fully described by auto- and cross correlation spectra of the field components. In case of digital circuitry clocked by a single clock pulse, the generated EMI is a cyclostationary process where the expectation values of the EMI are periodically time dependent according to the clock frequency and which have to be considered in modeling the EMI. Correlation analysis provides a basis for accurate modeling of noisy electromagnetic fields and, for strategies in computer aided design to reduce EMI. The determination of the near-field correlation spectra of the EMI radiated from devices and circuits by two-point measurements as well as methods to compute from these data the EMI distribution in complex environments are described. The amount of data required for the characterization of stochastic EM near fields can be reduced considerably by principal component analysis. The influence of noise on signal uncertainty is discussed.

This presentation will address the question of researchers, software developer or user of commercial software packages in microwave engineering: how can I modify the codes I am developing, or how can I utilize the software packages I am using in order to model uncertainties such as fabrication tolerances, tolerances of my input model, or statistical variations in the geometries I simulate? Naturally, it will start with a rigorous, yet intuitive introduction to fundamental aspects of the mathematical theory of uncertainty quantification in computational physics, from an engineering perspective. Then, the two families of uncertainty quantification techniques, namely non-intrusive and intrusive will be discussed. Along with Monte-Carlo, non-intrusive methods based on polynomial chaos and stochastic collocation driven surrogate models will be introduced. Intrusive techniques stemming from polynomial chaos will then be presented. The theoretical introduction of these techniques will be followed by examples, elucidating their application to microwave circuit analysis and design.

It is often the case that our electromagnetic simulations and measurement don't agree as well as we would hope. The origins for many of differences can be often traced to approximations or assumptions made during the simulation or measurement process (or both!). Such sources of uncertainty may include: approximations in numerical methods, imperfections in solid models, improper definition of electromagnetic port configurations, flawed material parameters, dynamic or inconsistent measurement environments, and potential drift or missteps in calibration. A detailed understanding of the electromagnetic fields, rather than just the port parameters, can aid in our understanding the origins of the differences. Measurement through electro-optic sampling of the dynamic electromagnetic fields above the circuit and their comparison with simulated fields often yields additional insight. Additionally, with Vector Network Analyzers now supporting uncertainty in addition to S-parameters measurement, we can now also begin to examine the contribution that uncertainty has in addressing the measurement-to-simulation gap…with sometimes surprising results.

Today’s RF & Microwave Modules continue to grow in both functionality and complexity. As such, uncertainties compound with the integration of multiple semiconductor, laminate, and component technologies. The importance and impact of accurate modeling for all building blocks must be considered, underscoring the importance of doing so with an integrated design and analysis platform. Once accurate models are secured and included for devices, packages, interconnects, laminates, components, and even evaluation boards…is that enough? Together we’ll investigate some key sources of uncertainty for Module design and explore how to mitigate risk with a comprehensive design, analysis, and verification flow.

Energy-efficient radio links are key to viable and widespread deployment of IoTs. Breakthroughs at the level of both circuits and systems, and energy harvesting and storage systems are much needed. In this talk, we will present a smorgasbord of the above mentioned problems and their potential solutions.

The Internet of Things (IoT) places new demands on wireless networks that are difficult to meet with conventional infrastructure, services, and protocols. A massive increase in the number of interconnected devices concurrent to IoT would strain the existing infrastructure, making some form of decentralized device-to-device or peer-to-peer (P2P) communication desirable. In order to manage the power consumption of P2P wireless links, duty-cycled radios have often been proposed due to their ability to shut off the static power consumption at low data rates. While earlier radio nodes for such systems have been proposed based on sleep-wake scheduling, such implementations are still power hungry due to large synchronization uncertainty (~1µs), and do not offer a scalable solution for large networks. A unique pulse-coupled oscillator based synchronization mechanism can be used to facilitate the network synchronization leading to both low power as well as scalable network realization. Pulse-coupled oscillator synchrony is modeled on a natural phenomenon observed in Southeast Asian fireflies, which are thought to use each other’s flashes to drive the network to flash in unison. This underlying theory can be used to synchronize radio networks, i.e. control the sleep-wake time of the radios, resulting into ultra-low power design. The talk will delve into the following: 1.) Duty-cycled Radios, and their effectiveness in low power applications. 2.) Timing and synchronization requirements for duty-cycled radios. 3.) Implementation of a scalable synchronization scheme based on naturally occurring phenomenon (fireflies blinking together). 4.) Design implementation of a UWB impulse based Radio, and consideration of pulse-shaping, FCC compliance, and spectrum usage. 5.) Interesting applications utilizing time sync methods. 6.) Implementation and Deployment challenges, and potential solution space. 7.) Requirements for future large scale and/or cellular compatible IoT systems

Meeting the demand for unprecedented connectivity in the era of internet-of-things (IoT) requires extremely energy efficient operation of IoT nodes to extend battery life. Managing the data traffic generated by trillions of such nodes also puts severe energy constraints on the data centers. Clock generators that are essential elements in these systems consume significant power and therefore must be optimized for low power and high performance. On one hand, sensor nodes require kHz clock generators operating within nW-scale power budget, while wireless and wireline transceivers need mW-scale GHz frequency synthesizers. In this talk, we will review the main design challenges of low power clock generators for both integer-N and fractional-N operation. We will then discuss recent design techniques such as injection locked clock multiplication to achieve both low jitter and low power consumption simultaneously.

Ultra-low-power sensing devices developed for next-generation IoT applications often communicate via deeply duty-cycled radios to achieve average system-level power consumption in the ~nanowatt regime. Absent the design of nanowatt wake-up radios, such sensing systems must keep synchronized via always-on oscillators that, unless carefully designed, can dominate the power in deeply duty-cycled systems. This presentation explores the design of on-chip integrated RC oscillators operating from Hz to kHz frequencies that, through architectural- and circuit-level innovations, achieve power consumption in the picowatt to nanowatt range. Various solutions to stabilize oscillation frequency from temperature and supply variations will also be covered.

The rapid proliferation of the Internet of Things (IoT) presents two seemingly disparate challenges of ultra-low energy overhead and high data rates to radio designers. But in reality, these two constraints are closely connected. Indeed, with the increase in the number of nodes in the mesh, the network becomes more and more congested. The only solution to decongest it, is to enable the radio to communicate at a higher data rate so that the system can be effectively duty-cycled. Now, in such a heavily duty-cycled system with high peak data rates, the average energy spent for communication becomes comparable with the energy overhead of the radio. In traditional radios employing synthesizers based on analog or digital PLLs, the energy overhead is dominated in by the energy required to wake-up the radio. The wake-up time is actually mostly determined by the very high Q of the crystal oscillator (XO), which at the same time also sets the phase noise performance. The wake-up time can be shortened without deteriorating the phase noise by building a loop-free frequency synthesizer based on an RF oscillator using a high-Q resonator such as a FBAR. But due to the frequency stability of the FBAR (after compensating for its frequency variations with temperature), tuning the synthesizer over a wide frequency range is extremely difficult. This talk will show different solutions to this problem and techniques for building low latency loop-free synthesizer.

Millimeter-wave (mm-Wave) technology is widely considered as one of the key technologies that will continue to serve the consumer demand for the increased wireless data capacity. The advanced CMOS can now well operate in mm-Wave bands, permitting the integration of a full transceiver in a low-cost, high-yield technology. However, due to the high operating frequency and large transistor size, the power amplifier (PA) becomes the most challenging building block in a mm-Wave transceiver. On top of that, efficiency enhancement techniques are still a hot topic in the design of RF PAs. In this talk, various PA efficiency enhancement techniques will be briefly reviewed and their implementations at RF frequencies will be compared. At the end, some state-of-the-art mm-Wave efficiency enhanced PA designs will be presented.

For next-generation 5G MMW applications, the 10-Gbps high-speed transmitters have stringing linearity and power saving requirements, which cannot be satisfied with the traditional digital pre-distortion (DPD) techniques. In addition, it is also required to work over a wide bandwidth, like from 57 GHz to 66 GHz. Furthermore, the design complexity needs to be taken into consideration. The RF pre-distortion linearization techniques aim to achieve adequate linear output power as PA operating under lower power consumption. Hence, this talk will cover the RF pre-distortion and post-distortion at 28GHz/ 38GHz/45GHz/60GHz bands for the future 5G MMW applications.

Doherty is a well known technique to extend the high efficiency region on PAs. In recent year, this technique has been implemented in Si and CMOS technologies, where the transmissionlines are replaced by a compact transformer-based power combiner. This allows the combination of Class AB and Class C biased amplifiers, exploits distortion cancellation and increases the output power. Furthermore, transformer-based combining and matching leads to higher efficiency than lumped-element based matching when using low-Q inductors. The result is an outstanding PA topology that gives an excellent trade-off between gain, efficiency, linearity and output power. This talk will give an overview of the different challenges when integrating a Doherty in CMOS at mm-wave frequencies and will discuss some examples in 40nm and 28nm CMOS.

Switching power amplifiers enable more efficient and higher power generation as compared with their linear counterparts, while they enable digital polar transmitter architectures. This talk covers various millimeter-wave switching power amplifier architectures where proper combination of single/stacked-transistor and harmonic-shaping passive components are judiciously used to ensure high voltage swing across the transistors (for high power generation) and small overlap between the voltage and current waveforms on each transistor (for high efficiency). Several proof-of-concept prototype switching power amplifiers at around 30 GHz, 45 GHz, and 90 GHz frequency bands realized in SiGe HBT processes will be covered

GaAs MESFETs (Metal Semiconductor Field Effect Transistors) and their later modifications HFETs (Heterostructure Field Effect Transistors) using artificially structured III-V semiconductors have been exhibiting the lowest noise temperatures of any microwave transistors since their first introduction in 1960’s. The current best noise performance is achieved by InP HFETs with gate as short as 35 nm. The rapid advances in technologies of HBTs (Microwave Heterostucture Bipolar Transistors) and CMOS pose a question whether these transistors can in the future offer alternatives to the extremely low noise performance of InP HFET’s (including their pseudomorphic and methamorphic versions). In order to compare these very different technologies noise models of unipolar and bipolar transistors are reviewed with emphasis on their common noise properties. Certain limits on the allowable values of transistor noise parameters are established proving that in practice any microwave transistor may be characterized only by three noise parameters, instead of four required for linear two ports in general. Experimental confirmations for III-V FETs, HEMTs, HBTs (in several different technologies including GaN HEMTs), and CMOS devices are shown. The question what determines the minimum noise temperature of either FET or BT and the current state of the art are briefly reviewed both at the room at the cryogenic temperatures. The possible limits of low noise performance upon further scaling of gate or emitter size are discussed. Widely held and widely published concepts in the treatment of noise in transistors and amplifiers, amongst those “gate induced noise” in FETs, wideband amplifier design, CMOS “noise cancelling” amplifiers are critically examined. The discussion is illustrated with experimental data

The measurement of the noise figure is at the core of the noise characterization of a linear microwave device and it is key to determine its noise parameters. Modern microwave test equipment have greatly enhanced and simplified the measurement of the noise figure in the past years - less so when considering the determination of the noise parameters. In either case, the measurement system is fundamentally tailored to measure 2 port devices and its adaptation to the characterization of modern differential amplifiers is not as straightforward as in the single ended case. Indeed, there are some clear limitations to the standard techniques in use when multiport networks are tested for noise figure or when transistors’ noise parameters are sought at high frequencies above the capabilities provided by commercially available tuners. This talk will start from a review of standard noise characterization procedures and expand to discuss 2 very recent accomplishments published by the author - the first on the determination of the noise parameters to linear N-port networks; the second on the noise parameters determination of on-wafer transistors without the use of tuners.

Noise in mixers continues to be a confusing subject, not only because there are at least three noise figure definitions for mixers, but also because the relationships between noise figure and noise temperature for two-ports are not valid for mixers. Since two-port noise concepts are not generally valid for mixers, the intuitive concepts we use to optimize noise in two-ports also are often invalid. In this presentation we will discuss the paradoxes of mixer noise characterization, show how well established noise characterization methods can be applied to mixers, and methods for optimizing mixer noise figure in both passive and active mixer circuits. We will also show some examples of modern, low-noise mixer circuits and provide some insight as to how they were optimized.

This talk addresses the question how nonlinear excitations impact the properties of flicker and white noise sources. In addition to frequency conversion due to mixing effects in nonlinear circuits, it has been shown that noise sources in semiconductors often depend on the large-signal current. Therefore, their noise spectra differ significantly from the linear case. This talk will briefly introduce the physical background of this effect, discuss approaches to simulate nonlinear noise in time and frequency domain, present a nonlinear noise model of a InGaP/GaAs HBT and conclude with examples of mixer and oscillator simulation and measurement.

Silicon photonics is a promising technology for realization of future data communications and interconnect systems. It brings the advantages of optical communication in a highly integrated platform with low manufacturing cost. In this talk we will cover Silicon-Photonic-based high-speed interconnects and their key building blocks. At the transmitter side design of compact modulators with low link-penalty as well as efficient CMOS drivers will be discussed. At the receiver side design approaches for high-sensitivity front-ends optimized for Si-P detectors will be presented. System-level topics such as WDM, 3D integration and packaging will be also covered. In particular effective co-design of electronics and photonics as a holistic approach for reducing the total power consumption and enhancing the performance of the link will be discussed.

Microwave photonics has been an active, but relatively small, research area over the past thirty years with applications to remote antennas. Conventional microwave photonic systems include optical modulators, lasers, photodiodes, and electronic drivers and result in performance trade-offs between bandwidth, linearity, noise figure, and power consumption. Additionally, size and area have been a significant consideration in the use microwave photonic components. The use of silicon photonic processes could offer lower cost and area and offer the ability to implement photonic integrated circuits but sacrifices some of the performance of spur-free dynamic range for a microwave photonic link. Consequently, this talk will address some techniques to augment microwave photonic links through the co-integration of electronics and photonics.

In this talk examples of electronic assisted photonics, where analog, RF, and mm-wave circuits and techniques are employed to improve the performance of photonic systems will be discussed. An optical synthesizer is presented which is capable of Hz-level tuning of a semiconductor laser emitting at 200THz over a 5THz range. Also, integrated electronic laser stabilization and phase noise reduction will be presented.

In this talk, we will investigate new devices emerging from a hybrid approach to the design of electronics and photonics integrated circuits. We will review several hybrid design design examples, such as optical phased arrays, hybrid equalizers, and optical mm-wave signal generation. Showing how the synergy among these technologies can results in more than some their parts.

Radio frequency phased arrays, primarily developed during the WWII for radar applications in discrete modular realizations, have entered consumer commercial applications such as automotive radars and high-speed wireless communications in monolithic realizations. Optical phased arrays enable several important applications such as lidar, imaging, display, and holography. This talk covers monolithic optical phased arrays realized in commercial SOI RF CMOS technology. The optical phased array architectures leverage lessons learned in the RF phased arrays and benefit tight integration of photonic and electronic functions.

This talk will describe what is expected from millimeter wave front-ends in the development of the future 5G networks. An overview of the most attractive frequency bands will be given, describing the related pros and cons. The current limitations and the foreseen evolution of electronics will be discussed.

We present several design techniques to achieve high efficiency and linear power amplifiers in the millimeter-wave frequencies. We will first review the performance of power amplifiers in different semiconductor process technologies at millimeter-wave frequencies. We will discuss the design, implementation and performance of stacked-FET power amplifiers, Doherty power amplifiers and linearization techniques to achieve high efficiency and linearity in millimeter-wave frequencies. The presented power amplifiers have applications in the 5G wireless communications.

The mm-wave frequency range is a key enabler for 5G wireless networks. Both for the backhaul network and access network with consumer products such as smartphones, tablets, cars, and IOTs. The 5G infrastructure is reaching out for higher frequencies having more available bandwidth. Simultaneously; price, handling and performance are important driving factors to meet the market requirements. The challenges for miniaturizing the MMICs and high frequency package solution reaching D-band are addressed and solutions to meet those demands are presented. Furthermore we will address the market and future outlook for the 5G mm-wave frequency bands.

This presentation deals with the challenges and possible solution for the integration, in SOI technology, of the power amplifier with the rest of the front-end.

The removal of circulators and isolators from high frequency front-ends has been discussed since ancient times, due to the high cost, size and weight of these bulk components that cannot be integrated. Unfortunately, the drawbacks from the removal of these crucial components are often too important, and the circulators and isolators stand still in the front-ends. This talk will show how, instead of removing them, circulators and isolators can be integrated and become convenient thanks to novel design techniques and advanced integration technology.

Outphasing architectures generate load modulation through phase control of multiple nonlinear PAs, offering the potential for linear amplification with high efficiency over a wide range of output powers. The advantage of this approach is in the efficiency of the branch PAs, which can be highly saturated. Historically, however, outphasing has several drawbacks that have limited its success. After an overview of historical outphasing techniques, this talk will present recent advances in outphasing PAs, including ones that reduce complexity and improve back-off efficiency in practical implementations. Limitations and challenges of designing outphasing PAs will also be discussed.

Pulsed load modulated (PLM) PAs can achieve high power efficiency through its duty-cycle dependent load behavior. Under certain Bitstream modulations, the amplifier output can be considered as a discretized and amplified version of the original RF signal. Filtering is thus required to eliminate the quantization noise and to restore the original RF signal. In this talk, we will introduce the different modulation techniques that may be applied to PLM PAs and filtering techniques including both passive and active filtering that work with the PAs to suppress the quantization noise without sacrificing the overall system efficiency.

A novel power amplifier architecture, the “Load Modulated Balanced Power Amplifier” (“LMBA”), is presented. The LMBA is able to modulate the impedance seen by a pair of RF power transistors in a quadrature balanced configuration, by varying the amplitude and phase of an external control signal. This enables power and efficiency to be optimized dynamically at specific power backoff levels and frequencies. Unlike the Doherty PA, the load seen by the active devices can be modulated upwards or downwards, both resistively and reactively, without any loss of power combination efficacy. The LMBA is presented as a potentially disruptive technique which enables any specific amplifier characteristic to be controlled dynamically over wide signal amplitude and frequency ranges. Implemented hardware examples will be shown, demonstrating LMBA action over octave bandwidths in S-band and X-band.

This paper shows the preliminary results of combining load modulation and envelope tracking, to get a large RF bandwidth with good efficiency in backoff. This Load Modulated Envelope Tracking (OLMET) combination method, is a Doherty like topology, without the bandwidth limiting quarter wave long lines. The technique is , in itself , independent of RF bandwidth. The only RF bandwidth limiting parts in this technique are the RF bandwidth of the amplifiers itself and a power splitter. Preliminary simulation results for a GaN MMIC process, show drain efficiency above 40% at 12dB backoff and a bandwidth of 1.2Ghz centred around 2.2Ghz.

A plethora of power amplifier architectures have been proposed to address the need for increased back-off efficency and linearity in wireless communication applications. Among them, varactor based dynamic load modulation (DLM) is one of the least investigated. This talk will first introduce the DLM concept. A variety of theoretical and experimental results will then be given to demonstrate the potential of DLM for both wideband, multi-band, and high power applications. We will also show how DLM in combination with dedicated digital pre-distortion linearization techniques results in a very competitive power amplifier architecture for both present and emerging wireless systems.

Most existing analyses of the Chrieix outphasing circuit assume that the active devices behave as voltage sources. Once this rusty creaking door is forced open, the analysis poses few problems and shows how the combined output power can be controlled over a useful range and with enhanced efficiency by varying the differential phase of the two input signals. But in almost all other applications and PA analyses transistors are not usually considered to behave as voltage sources, and as such it is surprising that after 80 years the outphasing configuration still sits on such shaky foundations. This paper analyses the Chireix outphasing circuit using a novel analytical model for the transistor I-V knee characteristics, rather than the approximation of a simple voltage source. It also incorporates input drive level variation, usually a critical independent variable in any other PA analysis, but curiously underrated by the RFPA outphasing community. The result is a more comprehensive understanding of how the outphasing circuit works, and various new design pointers.

This talk presents measurements of internal load modulation occurring in outphasing amplifiers with and without supply modulation. X-band GaN MMIC power amplifiers with 70% power-added efficiency and 2.7 W output power at 10.1 GHz are configured in hybrid outphasing circuits with several combiners that include bi-directional couplers, enabling calibrated measurements of internal load modulation. It is experimentally demonstrated that the load modulation critically depends on the power balance of the two internal MMIC PAs. Despite the additional loss in the combiner, peak total efficiencies greater than 47% are achieved by full outphasing PAs with more than 3.7 W of output power. A comparison between several outphasing configurations quantifies the improvement in efficiency for both isolated and non-isolated outphasing PAs with supply modulation.

With more integrated transmit and receive systems becoming increasingly prevalent, characterization tasks have received added attention. While calibration can be an issue, particularly with any over-the-air aspect of those measurements, linearity analysis can also be more involved. On the transmit side, added spectral content from clocking imperfections (usually appearing as spurs) or unintended coupling (both in an analog sense and in a digital-spectrum-intruding-on-analog-space meaning) can affect a spectrum-based linearity measurement and may require deconvolution. On the receive side, clocking imperfections can have a different effect on linearity measurements including noise floor elevations and low-order bit errors. This talk will look at some of the calibrations and interpretation nuances that can affect this class of measurements.

With digitally modulated radio signals increasing in carrier frequency and instantaneous bandwidth, the dynamic range demands of the signal chain in digital radios and test and measurement equipment is increasing. In order to transmit and receive signals with both small and large power levels, signal chains must utilize variable gain in order to maintain adequate dynamic range, usually measured by error vector magnitude (EVM). This workshop presentation will focus on tackling the tough problem of optimizing signal chains for digitally modulated signals with an emphasis on variable gain signal chains. In particular, the relationship between EVM and typical RF impairments such as noise figure, intermodulation distortion, phase noise, and linear amplitude/phase distortion will be discussed. Additionally, a methodology for optimizing variable gain signal chains will be shown. Finally, various graphical visualization methods will be developed to help signal chain engineers locate dynamic range limiting areas within the signal chain.

5G SDR approaches have an impact on nowadays microwave characterization technologies, the change in paradigm from analog to digital has a strong impact in the way nonlinear microwave characterization is seen. Behavioural characterization and modelling becomes a fundamental tool in complex systems, where the combination of analog and digital. In this talk a general overview of these technologies is presented, focusing on Software Defined Radio front-end characterization. Its characterization approaches will be presented as an integrated view on how to model and how to characterize those components from a behavioural point of view. Some examples on Analog to Digital Converters (ADC’s) and Digital to Analog Converters (DCA’s) as Digital Pre-distortion (DPD) feedback paths will be presented.

The use of digital predistortion (DPD) techniques for the linearization base station transmitter (BTS) power amplifiers is now commonplace in cellular wireless infrastructure. Digital predistortion is a classic example an adaptive control system, in which we are controlling the output of the plant – the power amplifier – using algorithms implemented in the digital domain, often using an FPGA or custom digital IC. It is a mixed-domain control system: the signal we wish to control is in an RF signal, and the controller is in digital hardware. Conversion of the RF signal to a digital signal, and back again, must be achieved without introducing further distortions and limitations. This places strict requirements on the data converters, frequency translation components, filters, and amplifiers that comprise the hardware aspect of the DPD system. Such considerations include the analog nonlinearity, IQ imbalance, and frequency response, and digital impairments such as jitter, thermal noise, and dynamic range.

High power amplifiers (PA), LNA and MIXERS have an important impact on RF front-ends behavior. Then, for designing a transceiver system, accurate behavioral models are necessary to take into account signal nonlinear distortion, bilaterality and memory effects. This paper presents an experimental measurement methodology and setup that allow a reliable identification of the Two-Path Memory Nonlinear Integral Model which is the basis of the description of each block of RF Front-ends. We discuss here the formulation of each models and the precision which can be obtained with a global macro-model (autonomous) which can be loaded in classical CAO tools (AWR, Ptolemy, Simulink ...). The study case presented here is a down-converter 3 GHz to 1.28 GHz the analysis bandwidth is 200 MHz.

There has been a recent renaissance of interest and investment in deploying high-data-rate communications networks based on constellations of 1000’s of Low-Earth-Orbit (LEO) satellites, as well as suborbital communications platforms such as High-Altitude Long Endurance (HALE) aircraft, persistent UAVs, airships, etc. These networks will have global impact on humanity by delivering ubiquitous high-bandwidth communications to nearly 60% of the world’s population that lives in underserved and fast-growing, but hard-to wire, regions of the world, maintaining such communications during natural or manmade disasters, with modest investments in ground infrastructure, and serving as a critical backbone for the Internet of Things (IoT). In the more distant future, these space-based networks may extend to serve manned and unmanned space missions throughout the solar system. We refer to these emerging networks as the Internet of Space (IoS). For example, Google and SpaceX recently announced a $B investment in a plan to deliver hundreds or thousands of micro satellites into LEO around the globe to serve Internet to rural and developing areas of the world. SpaceX’s ultimate goal is building a bridge to a future manned colony on Mars. Similarly, a new venture, OneWeb, is proposing a 648 satellite LEO constellation, with significant investments from Virgin Group and Qualcomm. Facebook and Google already have begun laying plans to serve under-wired markets with drone-based and balloon-based data networks. The European Space Agency and AirBus Defense & Space are planning a “Space Data Highway” that features EO satellites at GEO, and a set of LEO satellites to provide a hybrid optical / RF network for Emergency Response, Open Ocean Surveillance, UAS communication, Weather Forecasting and Wide-Area Monitoring on the impacts of human activities on state of natural resources (deforestation, loss of biodiversity, water/air pollution). Facebook is leading “Internet.org” to bring together technology leaders, nonprofits and local communities to connect the two thirds of the world that doesn’t have internet access. The Space-based networks represent the final frontier in the competition for connectivity. Back in the 1990’s, there were a number of large space-based satellite network ventures, such as Iridium, GlobalStar, Teledesic, etc. but only limited number of low-data rate (kbps) satellites were ultimately deployed. However, since that time, satellite technology has greatly advanced, bringing the cost of deployment down significantly. “Toaster-sized” micro-satellites can be launched dozens at a time to low earth orbits (LEO), reducing launch costs, while delivering performance comparable to larger, older satellites at higher orbits. Also, operation at LEO, satellites will also significantly reduce network latencies, while introducing challenging tracking, synchronization and handoff issues. Advances in microwave/mm-wave phased array technology and advanced CMOS over the last several years will also be key enablers. The new networks should not be expected to replace terrestrial networks, but will integrate seamlessly with these networks to provide ubiquitous global connectivity. Together with several other IEEE Societies the MTT-Society has launched an IEEE Internet of Space Future Directions Initiative in order to promote the future of satellite communication and sensing worldwide.

The presentation will review concepts for future satellite communication systems including LEO, MEO and GEO. Based on these future system architectures, the potential requirements for next generation microwave technologies will be derived.

Given the experience of space microwave equipment and commercial RF transceivers, combined with upcoming technology evolutions, transponder concepts are discussed and evaluated. The approach of total integration along signal path from input to power amplifier output enhances signal integrity and robustness in manufacturing. Due to full flexibility versus frequency conversion, a high reusability of this concept is supporting market expectation of individual frequency setting as well. Application of LTCC as core technology for 3 D integration of RF functions opens up the capability to merge amplifier, multiplier, mixer and bias circuit in a single module. The operating RF bandwidth selection at the in- and output is supported by tunable filters. Several concepts with tuning in orbit or on ground will be presented, based on different upcoming technologies. Tuning filters with Liquid Crystal to avoid moving parts or solutions featuring innovative coupling resonators to adjust filter bandwidth or manifold coupling and their contribution for small satellites are discussed and valued. Finally the presentation is summarizing different technology and design approaches for RF payloads, explaining their potential for space application and how they are fitting in low-cost transceivers for small satellite transponders.

Satellite technologies play an increasingly important role in a connected world, especially for advanced mobility applications. High data rate, abundant coverage, and ubiquitous positioning information are of major concern for many relevant use cases. Significant progress has been achieved in technologies both for the space segment and the ground segment, as well as for advanced testing methodologies that account for the intimate fusion of air interfaces and propagation channel. This contribution aims at providing a comprehensive and clear description of the potential of satellite payload and user terminal technologies for advanced mobility applications, accoun¬ting for the latest developments under the responsibility of the authors. The achievements include versatile electronically reconfigurable, space-qualified and tested, payload modules for the space segment in satellite communications. High-gain and low-profile tracking antennas and tracking mechanisms for heterogeneous satcom-on-the-move links represent major activities in the ground segment. The presentation further highlights the innovative testing technologies available in Ilmenau, covering a free-space testbed as well as virtual electromagnetic environments, for powerful over-the-air testing of mobile communication links. In a third part, approaches towards interference-resistant satellite navigation with compact antenna arrays for safety-critical applications will be described. This satellite-based technology has recently gained utmost relevance for automated driving.

The recent growth of low cost small satellite technology has fueled interest in in high data rate communications networks based on constellations of small satellites in Low Earth Orbit and opened new possibilities for interplanetary communications networks. Small satellites and CubeSats present a unique antenna design challenge due to the inherent stowage limitations, mass and environmental requirements. This workshop will present an overview of recent antenna technology developed to support the unique requirements of small satellites. Deployable high gain antennas and low gain proximity antenna technology will be highlighted. The workshop will also discuss future trends in antenna development for small satellite communications antennas.

The talk will present our latest work on SATCOM phased arrays using silicon technologies. It is seen that well designed chips can greatly lower the cost of SATCOM phased arrays, and can switch the beam very quickly (in microseconds). The use of these chips in actual implementations will be shown at Ku and Ka-band.

The internet of space technology has a lot to gain from programmable RF platforms. Technologies that result in fully reconfigurable transfer functions based on low-power mobile form-factor platforms are particularly attractive for satellite radios. It is for this reason that acoustic-wave resonators (BAW and SAW) need to be seriously considered for this field. However, conventional BAW and SAW architectures present few opportunities for achieving tunable responses. It is the purpose of this talk to discuss novel hybrid acoustic-wave lumped-element-resonator-based (AWLR) architectures that enable highly-reconfigurable operation. Such AWLR architectures bring the best of both worlds: a) the high quality factor of SAW/BAW resonators, and b) the wide tunability of lumped elements. Both bandpass and bandstop architectures for interference mitigation applications will be reviewed. Future opportunities for on-chip intrinsically-switchable filters will also be reviewed.

Emerging and future wireless solutions strongly rely on circuit agility to enhance equipment performance by improving form/weight factor and miniaturization. The introduction of reconfigurable and/or tunable circuits and antennas are considered as game changers in next generation of wireless communications (e.g. 5G) and detection (e.g. radar) systems, and more generically speaking for the internet of space. Endowed with high tunability and high quality factor, frequency agile components based in ferroelectric and MEMS technologies have been studied and optimized for a variety of applications going from filtering to phase shifting and from few GHz up to the mm-wave band. A perspective on recent achievements in this area will be presented.

Since the internet of space initiative requires a large number of satellites which need to be launched, a new technology is indispensable. On one hand, this technology needs to be tunable to provide a high flexibility in application but on the other hand it has to avoid mechanically moving components to reduce the risk of wear-out failures to a minimum. Furthermore, it should be low cost and adaptable to the higher frequency range, which might be an alternative for future satellite communication, due to the large available bandwidths. These challenges can be met by using electrically tunable systems such as 2D-steerable phased array antennas based on liquid crystals (LCs). Due to their unique property of exhibiting local anisotropy and their low dielectric loss in the higher microwave and millimeter wave range, it is a very promising material for the low-cost realization of continuously tunable RF devices for satellite communication. The technology has been improved in our research group over the past thirteen years, amongst others in projects funded by the German aerospace agency, for many different devices such as LC-filled hollow waveguide phase shifters for horn antenna arrays, microstrip line phase shifters integrated in planar phased array antennas or LTCC integrated antenna arrays. This work will therefore present the latest improvements of the LC technology for satellite applications.

In this presentation, a review of different approaches for the implementation of microwave components exhibiting multiband functionality, tuning and/or multifunctionality is addressed. All these approaches share the use of artificial transmission lines inspired by metamaterials. Examples of applications including passive components (filters, diplexers, splitters, couplers) and active devices (e.g. distributed amplifiers and mixers) are reported.

Impedance spectroscopy can yield important information regarding the electromagnetic response of biomolecules, cells, and other important fluidic systems. However, few impedance spectroscopy measurements are applied above the water relaxation frequency (approx. 18 GHz at 25C). We describe the design, verification and application of swept-frequency impedance spectroscopy experiments using microfluidic structures and wafer-probe measurements at frequencies above 20 GHz. Current broadband swept-frequency measurements are demonstrated over the frequency range 100 kHz - 110 GHz, while higher frequencies can be obtained only via band-limited approaches. Such measurements open up mm-wave frequencies for impedance spectroscopy investigations of biofluids.

The implementation of integrated mm-wave radiation sources and detectors offer a path toward a compact and low cost sensor for gas spectroscopy to meet the objective of a high-sensitivity, high-specificity breath sensor. The presentation reviews our recent work on transmitter (TX) and receiver (RX) circuits in SiGe BiCMOS technology for gas spectroscopy in the frequency ranges around 245 GHz and 500 GHz. The local oscillators of the TX and the RX are controlled by two external PLLs. The performance of our sensor system is demonstrated by using a gas absorption cell with dielectric lenses between the TX- and RX-modules, and measuring the high-resolution absorption spectrum of gaseous methanol (CH3OH) and acetonitrile (CH3CN).

Recent practical advancements in THz source, detector, and system technology have enabled researchers to explore a myriad of medical applications in both research laboratory and clinical settings. While much progress has been made in clinically relevant investigations, clinical translation has been limited. In vivo, physiologic tissue does not display specific THz frequency spectral signatures and features identified in the excised tissue are often difficult to observe in vivo due to the large aqueous background present in all tissues. Model based analysis has also been limited as THz frequency electromagnetic models are highly sensitive to tissue morphology. In practice, physiologic variation is so broad that it is nearly impossible to apply most models a priori. These issues are all compounded by biophotonic systems (optical imaging designs based on UV/VIS/IR spectra) that often offer similar or superior performance for a fraction of the cost and complexity. In light of these issues our group has chosen to focus on three applications which we believe are ideal for THz diagnostic imaging technology. The first two are acute burn wound severity assessment and surgical flap viability assessment. In both cases, the immediate physiologic response is characterized by massive changes in tissue water content (edema) allowing a THz based system to generate significant contrast without the need for substantial model based analysis or knowledge of tissue morphology. Further, excess tissue water content confounds the measurement of blood perfusion which is the key contrast mechanism of the majority of biophotonic systems thus suggesting a key advantage of THz frequency imaging. The third application is the early detection of corneal diseases that are correlated with changes in corneal tissue water content. Current practice limits diagnostics to the measurement of corneal thickness and extrapolation to water content which itself is incredibly inaccurate and does not account for physiologic variation. While disease related changes in tissue water content are small, the physiologic variation in corneal thickness, as referenced to THz wavelengths is nearly negligible. Thus the cornea can be treated as a thin film and model based analysis can be applied with reliable a priori assumptions on tissue morphology.

Global observations of clouds and precipitation are essential to improve prediction of severe weather having substantial impacts on human life and property. To this end, satellites in geostationary orbit (GEO) have greatly improved weather prediction by providing visible and infrared measurements on the 5- to 10-minute time scale. However, to peer inside of and help understand clouds, ice processes and the onset of precipitation requires millimeter-wave to THz sensors. At the same time, the satellite industry has recently experienced the maturation of disruptive technology and manufacturing processes to build and launch U-Class satellites. Commonly called CubeSats, these small satellites feature rapid development cycles (about two years) and low-cost launches (less than $0.5 M) as secondary payloads on missions of opportunity. Specifically, 6U-Class satellites provide healthy margins on mass, power, communications and antenna apertures to accommodate millimeter-wave to THz (90-900 GHz) sensors capable of observing clouds and precipitation on a global basis. A closely-spaced constellation of such remote sensors of precipitation can observe the time evolution of ice cloud processes leading to precipitation with revisit times on the order of 5 minutes. Currently, a 6U-Class satellite is being produced to perform technology demonstration for a constellation mission known as the Temporal Experiment for Storms and Tropical Systems (TEMPEST). A partnership among Colorado State University (lead institution), NASA/Caltech Jet Propulsion Laboratory and Blue Canyon Technologies, TEMPEST-D is planned to be delivered by July 2017 for a NASA-provided launch by April 2018.

We describe the integration of microfluidics with high quality factor electrical resonators for the creation of lab-on-a-chip devices for the analysis of small quantities of biological, toxic, explosive, and other liquid types at terahertz and microwave frequencies. These devices are capable of measuring the complex permittivity of sub-microliter liquid samples, and we demonstrate this sensitivity by detecting individual biological cells in a free-flowing buffer solution. The dielectric measurement of single biological cells in the terahertz and microwave bands represents a possible route for the label-free detection of circulating tumour cells in blood samples, and we present results towards realizing this goal.

In spite of the considerable progress in terahertz technology, practical feasibility of many exciting applications of terahertz systems is still bound by the low power, poor efficiency, and bulky nature of existing terahertz sources. Photoconduction is one of the most promising and commonly used means of terahertz generation, due to availability of high power, wavelength tunable, and compact optical sources with pulsed and continuous-wave operation required for broadband and narrowband terahertz generation, respectively. Here, we present an overview of recent advances in photoconductive terahertz emitters that utilize plasmonic nanostructures to significantly enhance optical-to-terahertz power conversion efficiency by enhancing light-matter interaction at nanoscale. Utilizing plasmonic nanostructures in a photoconductive emitter allows concentrating a larger fraction of the incident pump photons within nanoscale distances from the contact electrodes. By reducing the average transport path of photocarriers to the contact electrodes, the ultrafast photocurrent that drives the terahertz antenna is significantly enhanced and the optical-to-terahertz power conversion efficiency is increased considerably. This enhancement mechanism has been widely used in various photoconductive terahertz emitters with a variety of device architectures and in various operational settings, demonstrating significant optical-to-terahertz conversion efficiency enhancements. We demonstrate that use of two-dimensional and three-dimensional plasmonic nanostructures leads to 2 orders-of-magnitude and 3 orders-of-magnitude enhancement in the optical to terahertz conversion efficiency of conventional photoconductive emitters, respectively, offering a record-high optical to terahertz conversion efficiency of ~10%. We show that the significant performance enhancement offered by plasmonic nanostructures can be utilized to achieve record-high terahertz power levels in both continuous-wave and pulsed operation at optical pump wavelengths ranging from 800-1550 nm.

A global review of the actual state of the art of THz Sensor activities is given. This includes biomedical and environmental applications as well as industrial testing. In this presentation the focus is put on the electromagnetic/physical aspects not on the signal and image processing which is an important part in some applications as well.

Increasing optical and electrical functionality within a volume drive the continued sub-wavelength scaling of nano-scaled devices. Studies of light matter interaction (LMI) in the sub-wavelength regime have revealed unique absorption, scattering, and transmission characteristics. Unconventional devices based on such characteristics include sub-wavelength perfect absorbing films, devices with tailored scattering signatures, and devices with electromagnetically induced transmission, to name a few. To achieve smaller volumes and enhanced LMI, extreme light confinement is needed and is realized with plasmons, which are sub-wavelength surface electromagnetic waves at an interface with permittivities of differing sign. Recent research in sub-wavelength plasmon metasurfaces offer significant potential to control far-field light propagation through the engineering of amplitude, polarization, and phase at an interface. We discuss in this presentation demonstrated phase modulation from an electronically reconfigurable metasurface based on a van der Waals material, graphene. Based on experimental data, we discuss the feasibility of reconfigurable mid-infrared beam steering devices based on a 2-D material.

This talk will present an overview of the European project PlanarCal, which is funded from the European Metrology Programme for Innovation and Research (EMPIR). The overall aim of the project is to enable the traceable measurement and electrical characterisation of integrated planar circuits and components from radio-frequency (RF) to sub-mm frequencies. This will allow industry to characterise components and devices for eventual use in high-speed and microwave applications (e.g. wireless communications, automotive radar and medical sensing) with known measurement uncertainties.

The National Institute of Standards and Technology has put together a traceability path for large-signal on-wafer measurements that starts with on-wafer measurements and extends all the way through circuit design and simulation. We will discuss this entire traceability path. We will begin with the fundamental linear part of the calibration, which focuses on on-wafer impedance and scattering-parameter measurements. Then we will turn to the amplitude and phase aspects of the calibrations. We will discuss how these uncertainties can be propagated through the device modeling process, and how we can add process variations to the results. Finally, we will discuss the creation of models that capture not only the impact of measurement uncertainty and process variation in the model development process, but touch on how these can be seamlessly implemented in ADS and other circuit simulation tools.

Depending on the application there are a multitude of calibration algorithms (SOLT, TRL, LRL, LRM, …) which are applied to on-wafer measurements. Additionally there are many uncertainty contributions (contact repeatability, drift of the VNA, cross-talk, uncertainty of the standards, ...) which have to be considered for estimating measurement uncertainty. VNA Tools is a free software which can be used to calibrate on-wafer measurement data and which propagates the uncertainties to the final result. Starting from the experimental characterization of uncertainty sources up to the final result with error budget, all steps are shown in an exemplary on-wafer calibration.

On-wafer probing technology have widely been demanded at millimeter-wave and Terahertz frequencies. However in order to make precision and reproducible measurements, all other aspects of the measurement techniques must be considered. This talk will present the AIST’s probe contact algorism and over-determined wafer-level calibration technique in order to precision and reproducible wafer-level VNA measurements that can be achieved.

Wafer-level S-parameter measurement at mm-wave and sub-mm wave frequencies plays a crucial role in the model development and IC design verification and debug of advanced semiconductor technologies. Accurate calibration of the entire wafer-level measurement system to the RF probe tip end or to the intrinsic device terminals is a critical success factor for extracting trustable device model parameters and characterizing true performance of a MMIC. This presentation will start with the basics of S-parameter measurement and calibration techniques at the wafer-level. Special attention will be paid to how to choose the right calibration method for specific measurement application needs. Definition of the calibration reference plane and the measurement reference impedance of a calibrated system will be reviewed as well. Finally, the potential sources of calibration residual errors will be analyzed. Practical examples will be given on how to minimize the impact of such errors on the measurement accuracy of a calibrated probe system.

The ongoing miniaturization of electronic devices has led to the discovery of new nanomaterials and new phenomena at the nanoscale. In turn, this has led to the design, fabrication, and development of RF nanoelectronic devices that incorporate nanoscale elements or nanomaterials, such as carbon nanotubes, semiconducting nanowires, or graphene. Reliable, accurate, on-wafer measurements of such devices are critical to their optimization and commercialization. To this end, a full framework, including measurement, modeling, and validation, has been developed for on-wafer characterization of RF nanoelectronics. The calibration approach is based on the on-wafer, multiline thru-reflect-line technique. Further, this framework addresses the inherent impedance mismatch between RF nanoelectronic devices and commercial test equipment. Finally, circuit and finite-element models are used to extract circuit and material parameters for the devices.

Nanotechnology emerges from the physical, chemical, biological and engineering sciences, where novel tools and techniques are developed to probe and manipulate single atoms and molecules. In particular, the introduction of near-field scanning microwave microscopy tools have pioneered many applications, notably including mapping and quantitative measurement of complex impedances of nano-devices and electromagnetic properties of materials. In this frame, a unique scanning 1-110 GHz scanning microwave microscope built inside a scanning electron microscope is developed. The system can produces simultaneously complex impedance, atomic force microscopy and scanning electron microscopy images providing novel and unique equipment featuring unprecedented capabilities for tackling the frontiers between spatial resolution and frequency domain.

In the last few years electronic devices increased their corner frequencies tremendously, on the one hand due to material optimizations and on the other hand due to the ongoing miniaturization in device size. Especially the miniaturization of device size resulting in an increase of the corner frequencies leads to parasitic effects like fringing or coupling that have an influence on the device behavior even at lower frequencies and therefore have to be well described in the modelling process. In this talk we will show, that specifically the stability prediction of devices models, even at low frequencies, can be much improved, when we consider the above stated high frequency effects. Nevertheless the precise characterization of these effects needs reliable on-wafer measurements at sub-millimeter frequencies, which is of course not straight forward achievable with a classical design of the on-wafer access- and test structures. Therefore this talk will additionally look at obstacles that occur during the planar on-wafer measurement for the device characterization at sub-millimeter frequencies. At the end we will present some interesting new approaches that are going to improve the main problems of device characterization at sub-millimeter frequencies.

With the continuous up-scaling of the maximum operation frequency of commercially available integration technologies, mm-wave circuits are entering real-life applications, such as automotive radar and high data rate wireless and wired links. In order to foster these device improvements and increase the penetration of mm-wave applications in the commercial world, the availability of accurate measurement techniques, for low-cost, large-volume technology platforms, is becoming a key requirement. However, the need to design accurate calibration kits in the same medium as that of the DUT, due to the error arising from calibration transfer at mm-waves, is colliding with complexities and stringent design rules encountered when integrating components is silicon technologies. In this presentation, an overview of all currently employed techniques for defining the parameters required by TRL calibrations are presented, highlighting advantages and drawbacks of the available approaches. Design guides to implement high precision TRL kits up to the sub-mm-wave range are given. The main drawbacks in terms of propagation of un-wanted modes and improper coupling through the probe to pad transition are also analyzed. Finally, the evaluation of the quality achieved by calibration kits integrated on commercially available silicon technology is presented.

This talk covers essential aspects of modeling surface roughness for microwave applications based on underlying physics. At first, surface roughness metrology and commonly used roughness parameters are described. Existing models and their limitations are discussed before the recently proposed Gradient Model is introduced. To this purpose, the modeling approach, the derivation from Maxwell‘s equations, model predictions and their experimental verification are shown. Then a corresponding surface impedance concept is derived, which allows for easy application of the Gradient Model with 3D field solvers or analytical models. Therewith obtained simulation results illustrate roughness impact on loss and phase delay in typical transmission lines. Comparison to measurement results up to 100GHz show, that the Gradient Model accurately predicts these quantities for rough conductor surfaces. Furthermore the impact from imperfect surfaces on planar CPW calibration standards is shown.

Taking together the NonLinear Vector Network Analyzer and the Spectrum Analyzer capabilities of a modern network analyzer makes available very accurate vector (IQ) measurements of wideband modulated test signals. As the active DUT (can be power amplifier or frequency converter devices) input and output waves are measured coherently within a calibrated network environment, new insights exhibiting the DUT stimulus/response under modulated signals are demonstrated.

The talk focuses on the recently introduced dynamic-bias measurement technique for transistor characterization at microwave and mm-wave frequencies. We will discuss various measurement set-ups to perform dynamic-bias measurements and show how to use these measurements in the modeling phase. Furthermore we will introduce dynamic-bias S-parameters, which can be directly derived from dynamic-bias measurements and represent a natural extension of classical multi-bias S-parameters. In particular, we will show how these parameters allows one to obtain a more effective characterization of low- and high-frequency dispersive effects.

Receiver performance can sometimes be a limiting issue in mm-wave, wide modulation bandwidth systems. The higher and wider IF frequencies can present some unique linearity and correction challenges particularly in variable gain and A/D conversion areas. This talk will explore some of the more subtle linearity and distortion issues at multi-GHz IFs, how they can be characterized, and sometimes mitigated or corrected.

The content might be described as "Under large-signal modulated excitations, RF PAs show significant inter-modulation and harmonic nonlinearities at the same time. In order to characterize the actual PA behavior in real life, the multi-harmonic modulated PA input/output waves have to be entirely and correctly measured. In the presentation, newly proposed phase reference and phase calibration techniques are introduced and discussed, based on which the NVNA test-bed could be developed for the full characterization of RF PAs under various wideband excitations. Moreover, other potential techniques and non-mature ideas under development are shared for discussion.

The behavioral models used for the representation of CW and modulated multi-harmonic data will be reviewed in this lecture. This will include the general multi-harmonic Volterra functions for CW periodic nonlinear RF excitations, the X-parameter/S-function approximations for mildly nonlinear RF excitations and their extension for broadband modulated multi-harmonic signals. Next this lecture will consider the characterization and mitigation of the impairments associated with the PA nonlinearities in SDR systems when the same power amplifier is used for the amplification of concurrent multiple-band signals (carrier aggregation). Both predistortion and feedforward approaches for modulated harmonic cancelation will also be presented. Finally this review will conclude with a discussion on nonlinear impairments in MIMO systems and advanced configurations for self-testing and adaptation which might be called up on for their mitigation.

We present a digital pre-distortion (DPD) method based on dynamic X-parameters. We first explain how long term memory effects can be modelled by dynamic X-parameters. Next we show how a dynamic X-parameter model can be inverted in order to generate a robust DPD method. The resulting DPD is more robust than existing DPD techniques as it works well for a wide range of modulation bandwidths and signal amplitude distributions.

The continuous growth of data rate requirements in modern wireless communications leads to more-and-more complex and wideband modulation signals that need to be processed by the transmit power amplifier with high fidelity at the lowest power consumption. These are making design, characterization and modeling of the power amplifier very challenging due to the combination of wide variability of the signal time statistics, high dynamic range and very large bandwidth. The presentation will summarize some recent advances on the behavioral modeling methodologies of the nonlinear memory of power amplifiers.

Massive MIMO, MMIMO, systems for 5-G wireless networks pose new problems to nonlinear transmitter modeling and its extraction. Contrary to conventional transmitters where the output amplifier drives a fixed load, mutual coupling between the elements of the MMIMO active antenna array is seen by the interacting power amplifiers, PAs, as a form of dynamic load modulation. Therefore, the fixed load condition, which is one of the strongest underlying assumptions in the traditional low-pass equivalent transmitter behavioral models, can no longer hold, and a single-input/dual-output formulation, for the nonlinear dynamic PAs, followed by a multi-input/multi-output network, representing the linear and dynamic radiation sub-system, must be adopted. The present talk addresses this new transmitter behavioral model formulation and the nonlinear vector network analyzer systems required to extract them. Various possible behavioral model formulations are reviewed in terms of complexity and accuracy while their corresponding measurements systems will be discussed in terms of both the required hardware measurement set-ups and modulated stimuli.

In this talk mixed-signal instrumentation and measurement approaches will be presented to characterize software defined radio and digital radio front ends for the new 5G communication paradigm. The talk will present the main drawbacks, the calibration procedures and the framework to apply D-parameters to mixed-signal integrated solutions for 5G. Some practical examples will be showed including the application of this approach to digital pre-distortion approaches.

In the early 90's , Cellular telecommunications adopted the GSM ( 2G) standard and opened up a significant market segment for GaAs solutions in the form of 4W GaAs HBT PAs for the handsets and low noise receivers for the base stations.Eversince , Si solutions have been catching up and nibbling off at the GaAs MMIC market share. The upcoming advent of 5G mobile telecommunications, will again put III/V solutions in the forefront. The move to Ka band hand sets as well as Ka , E ,and W bands infrastructure, paves the way to very advanced and economically sustainable GaN/Si transmit and receive MMICs. In this context , OMMIC has developed 100nm and 60nm GaN/Si processes offering unequalled power , linearity and low noise performance up to 100GHz at 12V, meeting the system requirements and fully replacing GaAs solutions. A full range of relevant 5G MMICs will be presented at the workshop including Ka band 10W PAs with 30% PAE and a full T/R chip.

In this presentation, various techniques to implement linear power amplifiers (PAs) for application in 5G communication will be first introduced. Specifically, techniques that are used to design SiGe and CMOS SOI RF to mm-wave PAs will be discussed in details. A technique adopted by the presenter and his group that is based on stacking of power amplifier cells in scaled CMOS SOI technologies is also presented. Stacking of PA cells facilitates relatively high output powers and wide bandwidths without a need to utilize on-chip power combining networks. Mechanisms that degrade linear output power and efficiency of PAs are also identified and ways to suppress these effects and enhance the PA power performance are presented. Several examples of high efficiency linear PAs based on standard scaled SiGe BiCMOS and CMOS SOI technologies with output powers approaching 1Watt and peak power added efficiencies in the excess of 20% will also be presented. Possible future directions for SiGe and CMOS SOI PAs for 5G applications and beyond will also be discussed.

Short Description: This paper will review the challenges for 5G mmWave power amplifiers for mobile and basestation terminals and examine a range of PA architectures, efficiency enhancement and linearization schemes. From the viewpoint of system and practical implementation requirements, the most promising techniques will be highlighted and compared in terms of the key figures of merit.

Mobile and backhaul transmitters for high-capacity networks are increasingly concerned with power consumption constraints. The desire for high average efficiency in RF and millimeter-wave systems has spurred interest in load and envelope modulation circuit techniques that are compatible with CMOS and/or SiGe BiCMOS technologies. This talk will present challenges confronting for wideband (up to 2 GHz) and high PAPR waveforms that have been proposed for 5G systems. I will present several examples of circuit solutions that we have developed to improve PA performance at back-off operating conditions through the use of Doherty, outphasing, and envelope tracking techniques at both microwave and millimeter-wave bands.

Compound semiconductor technology, and specifically GaAs, has captured a large and growing market share in wireless and optical systems by providing the optimum combination of RF performance and value. To remain the solution of choice in next generation systems and applications, GaAs technology has to compliment its inherent performance advantage with increased integration and functionality. Historically, GaAs has lagged Si technology in offering multiple device types on the same wafer (e,g, power, low noise, E/D logic, schottky diode, PIN diode, etc) to enable highly integrated multifunctional MMICs. This gap is rapidly closing and this presentation will describe several advanced GaAs platforms that incorporate new levels of functionality within high performance GaAs HBT and pHEMT technologies. These platforms provide users with a new set of tools to address the ever evolving and complex performance requirements of present and future communication systems.

From a PA performance point of view GaN PAs offer a lot of efficiency enhancement for 5G applications, especially in the new targeted PA bands from 3-6 GHz and at the mm-wave. Gallium Nitride IC technology further justify their existence in the 5G by very good performances for backhaul links around 30 GHz, around 60 GHz and at E-band. Further the good performance allows reduction of transistor numbers in amplifier stages which allows compact module integration in wavelength-limited spacing of active MIMO antennae to be deployed. The paper discusses novel Pas and MMICs between 3-6 GHz, 30 GHz, 60 GHz to 84 GHz which are in-line with the recent performance advancements of Gallium Nitride while maintaining the cost balance.

He will talk about a PA line-up consisting of CMOS based pre-PA and GaN power stage for array-antenna systems. He would also explain the biasing and power-on sequencing

This session will discuss the challenges and trade-offs for millimeter-wave 5G systems and the specific impact to PA requirements. Starting from a basic understanding of the 5G link budget and phased array architecture, the derived per element PA power levels are presented. Then an analysis on total power consumption will highlight sweet spots based on the performance and efficiency of different device technologies. The main frequency bands, spectral masks, and ACPR requirements will be discussed. A system level transmit chain line-up analysis will highlight the gain and linearity budget needed to support the high-order multi-carrier modulation schemes being proposed. Lattice spacing requirements and constraints on die size, packaging, and minimum integration levels will be discussed as well as thermal challenges and considerations. Finally, the advantages of GaN PAs for base-stations is discussed and recent performance results are presented.

Efficient non-contact vital sign accurate detection methods are needed to develop continuous tracking of elderly population, infants, and suicidal subjects 24/7 observation. Doppler radars and optical based imaging photoplethysmography (IPPG) have been successfully used non-contact vital sign detection. Both methods are presented and their limitations will be discussed.

THE PROBLEM? Measuring and monitoring any patient's vital signs is essential for their care- in fact, all care first begins by collecting vital signs like heart rate and blood pressure. The current standard of care is based on monitoring devices that require contact - electrocardiograms, pulse-oximeter, blood pressure cuffs, and chest straps. However, contact-based methods have serious limitations for monitoring vital signs of neonates as they have extremely sensitive skin and most contact-based vital sign monitoring techniques result in skin abrasions, peeling and damage every time the leads or patches are removed. This results in potentially dangerous sites for infection increasing the mortality risk to the neonates. OUR SOLUTION? We propose to use normal camera to measure the vital signs of a patient by simply recording video of their face in a non-contact manner. From the recorded video of the face, our algorithm, distancePPG, extracts pulse rate (PR), pulse rate variability (PRV) and breathing rate (BR). The algorithm is based on estimating tiny changes in skin color due to changes in blood volume underneath the skin surface (these changes are invisible to the naked eye, but can be captured by a camera). Our algorithm, distancePPG (patent pending), achieves clinical grade accuracy for all skin tones, under low light conditions and can account for natural motion of subjects. It does so by intelligently combining skin color change signal from different regions of the visible skin in a manner that improves the overall signal strength. Our algorithm results in as much as 6dB of SNR improvement in harsh scenarios, rapidly expanding the scope, viability, reach and utility of CameraVitals as a replacement for traditional contact-based vital sign monitor.

Atrial Fibrillation (AF) is the most common sustained arrhythmia. Over 5.2 million Americans have been diagnosed with AF, and the prevalence of AF is increasing concomitant with the aging of the U.S. population. AF exerts a significant negative impact on the longevity and quality of life of a growing number of Americans, predominantly through its association with an increased risk for heart failure and stroke. Effective AF treatments reduce a risk of complications from AF. A major challenge facing clinicians and researchers is the early detection of AF, because particularly in its early stages, AF can be intermittent and asymptomatic. While the population with undiagnosed AF is substantial, studies have shown that more intensive cardiac monitoring can improve AF detection and enable timelier institution of treatment. Automated AF detection algorithms offer real-time realizable AF detection but often suffer from the fact that common benign causes of rhythm irregularity, most notably premature atrial (PAC) and ventricular (PVC) contractions, can cause false positive AF detection. There is a pressing need to develop a continuous arrhythmia monitoring device that can accurately and reproducibly distinguish between AF, NSR, and premature beats (PACs and PVCs) in order to improve patients’ cardiovascular health and reduce the costs associated with treating AF. To this end, we have recently developed a smartphone application for arrhythmia discrimination, which can identify NSR, AF, PACs and PVCs using pulsatile time series collected from a smartphone’s camera. This application detects and removes motion and noise artifacts (MNAs). This talk discusses the development and clinical testing of arrhythmia discrimination. Given the ever-growing popularity of wearable devices and smartphones, our approach to arrhythmia discrimination will give the population as well as health care providers the opportunity to monitor arrhythmia under a wide variety of conditions outside of the physician’s office.

As a crucial advantage, the self-injection-locked (SIL) radar is highly sensitive and inherently immune to stationary clutter, such as that produced by background reflection and antenna coupling. This work presents the superiority of the SIL radar over the conventional continuous-wave (CW) radar in terms of sensitivity under the same conditions of clutter and power consumption. Moreover, with the help of signal processing or mutual-injection-locking techniques, an advanced SIL radar was developed to monitor human vital signs with reduced body motion artifacts.

In this talk, we will present test results of of monitoring respiration and heartbeat of laboratory rats/mice using a 60-GHz system-in-package integrated micro-radar to perform non-contact, non-invasive measurement. The system hardware, detection method, and data processing algorithm will be described. This system allows the lab animals’ respiration and heartbeat to be monitored continuously without the need of surgical implants and will eliminate the fatality rate through the surgical/recovery process

Body worn sensors can be used to capture a wide variety of human biometric data, and with the appropriate wireless infrastructure this information can be used to empower disruptive new systems for healthcare, security, and human machine interaction. Lightweight unobtrusive sensors worn close to the skin of as attachments or clothing can measure cardio-electric signals, bio-impedance, cardiopulmonary and limb motion, gestures, activity, and other useful biometric quantities. Such sensors can be passive or active, and in some cases can also harvest energy associated with these measures to power both sensing and wireless communications functions. The information collected by such systems can be used for applications including medical analysis for healthcare tracking, motion capture for motion compensation or virtual and augmented reality gaming or simulations, and for applications in subject identification and security. This presentation will cover theory and techniques for sensors, communications, and applications for a variety of wireless wearable systems.

Wearable smart sensors with embedded control and communication links have the potential to improve the quality of service in healthcare, security monitoring, and energy conservation. This presentation provides an overview of our research activities on wearable radar sensors for indoor tracking and health monitoring applications. In a smart building, short-range radars can map the environment and monitor health condition to benefit the human well-being. Starting from basic motion and range measurement theory, our recent research efforts on wearable FMCW/interferometry/UWB radar for position tracking, fall prevention and detection, and localization will be discussed. Technical details will be presented, followed by video demonstrations of the developed technologies. Finally, the outlook of wearable radar sensor will be discussed.

The medical device industry has been seeing a significant jump in technology with notable advancements in size and wireless capabilities. Increasing competition and maturing markets, particularly in cardiology applications, are also drivers for the recent advancements. Device miniaturization is allowing for pacemakers to be implanted via a catheterization procedure while some diagnostic devices can now be done in a simple doctor’s office procedure. On the wireless side, medical technology has avoided using the same frequencies as the consumer market due to concerns including interference and security. More recently, however, implantable devices are seeing a trend towards compatibility with consumer market devices with the use of bands like Bluetooth and the concept of bring-your-own-devices.

Imaging photoplethysmography (IPPG) is noncontact physiological signal detection technology based on the traditional photoplethysmography (PPG), which could achieve some specific cases of clinical and daily detection such as physiological signal detection with open wounds and motion state. IPPG technology with its non-contact measurement, low cost and easy operation, has become one research hotspot in the field of the instrument and biomedical engineering. At present, some important physiological parameters such as heart rate, respiration rate, oxygen saturation, heart rate variability, blood pressure etc, have been detected through IPPG technology by our research group.

Abstract: Emerging 5G standards will actively rely on silicon mmWave systems as an enabling technology. This talk will provide a brief system context for the use of silicon in mmW applications, with particular attention paid to technology tradeoffs.

The increasing device performance of scaled bulk CMOS technologies (approaching 300GHz range for Ft and Fmax) has resulted in a wide usage of RFCMOS technologies, e.g. for cellular and connectivity applications. RFCMOS technologies are differentiated by enhanced device modeling, additional passive devices and improved metal stacks compared to their digital counterparts, but the device performance is typically a result from digital scaling. To cover additionally mmWave applications, it is important to understand how digital performance scaling (which is the driver for CMOS technologies) correlates to mmWave device performance. This correlation will be shown, based on scaling of physical device models. Based on this understanding and the CMOS technology roadmap and features, the best CMOS technologies can be selected. Trends and possible technology enhancements will also be discussed.

This talk presents a simplified version of the charge-based BSIM6-EKV MOSFET compact model and shows how it can be used to assess different bulk CMOS technologies in terms of current efficiency and RF performance using just a few parameters. The concept of inversion coefficient IC is first introduced as an essential design parameter that spans the entire range of operating points from weak via moderate to strong inversion. It is then shown how several important figures-of-merit (FoM) including the current efficiency G_m/I_D , the transit frequency F_t, their product (G_m·F_t)/I_D and the minimum noise factor F_min can be expressed in terms of IC to capture the various trade-offs encountered in RF circuit design. It is then explained how short-channel effects, and mainly velocity saturation, have significantly slowed down the increase of F_t and degraded F_min in recent technology nodes. The simplicity of this IC-based model is emphasized by comparing it against measurements from a 40- and a 28-nm bulk CMOS processes and BSIM6 simulations. Finally, it is shown how this simplified model can be extended to FDSOI and FinFET and used for a fair comparison between bulk, FDSOI and FinFET.

5G is the new era of networking that will expand upon the human possibilities of the connected world. Radio components that are both scalable (i.e., for mMIMO) and flexible (i.e., device integration) are key catalysts in realising low-latency, high-throughput networks. RF CMOS device scaling has provided a means to address cost-effective solutions, however, foreseeable 5G system requirements demand more beyond CMOS lateral scaling techniques. In this presentation, mm-wave RF CMOS devices, circuits and architectures will be examined, with an emphasis towards potential 5G systems applicability. Opportunities for technology enhancements will be discussed, while disseminating circuit design techniques that are amenable for the mm-wave connected world.

Advanced SiGe BiCMOS technologies offer today high-frequency SiGe HBTs with cut-off frequencies fT and fMAX beyond 300 GHz addressing a wide range of RF and mm-wave applications including high-data rate wired and wireless communications and radar for autonomous driving. Research results have demonstrated the potential for substantial further performance improvements of SiGe HBTs in future technology generations. SiGe HBTs with fT values of 500 GHz and fMAX values 700 GHz were realized recently. This talk will address performance perspectives and challenges for next generation SiGe BiCMOS technologies which integrate such HBTs with advanced CMOS nodes. We will discuss implications for existing applications in the 25, 60, and 77 GHz frequency bands and potential new application areas up to the lower THz range.

The emergence of mmWave applications, like 5G and Satcom, has opened up opportunities for the high-performance SiGe BiCMOS technologies which provides a sweet spot for performance and integration. In this talk we will present the GLOBALFOUNDRIES 130nm / 90nm SiGe BiCMOS technology portfolio for building mmWave front-end blocks cost effectively and give exemplary circuit examples. We will look at the challenges for future bipolar scaling to address the needs for these mmWave applications.

This tuturial will start with a brief overview on the most relevant physical effects in high-speed SiGe HBTs. The associated compact formulations and their integration in an advanced compact model will be presented. Next, the procedure for an as independent as possible determination of the model parameters will discussed with emphasis on being able to generate physics-based geometry scalable large-signal compact models. Selected examples for extraction steps and related results as well as for a comparison between model and experimental data of SiGe HBTs fabricated in the most advanced process technology will be provided. In particular, combining a physics-based compact model with a geometry scalable parameter extraction enables a quantification of the impact of various physical effects on device performance as valuable feedback for process development. Further model verification will be demonstrated by comparing simulation and measured data of various mm-wave benchmark circuits. Finally, major issues faced in the future regarding reliable compact modeling, parameter extraction and measurement capability will be discussed.

Experimental results for RF and wideband benchmarking circuits fabricated in 0.13um and 90nm SiGe-BiCMOS technologies are presented. The circuits include: a 130 GHz-bandwidth feedback amplifier, 54-65 GHz power amplifier, DC-100GHz frequency multipliers and dividers, and wideband data modulators. Technology aspects related to passive and active components on-chip applicable to all silicon technologies are also described.

This talk will highlight circuit and system capabilities for SiGe BiCMOS technology, for two specific case studies. First a W-band radar transceiver will be presented, and the key building blocks of the power amplifier, voltage-controlled oscillator, and mixer will be evaluated, together with technology aspects which result in improved performance. Second, millimeter-wave power amplifiers and transmit beamformers will be highlighted, together with design techniques which take advantage of the bipolar transistor to achieve high output power and high efficiency.

SOI technology is a great platform for RF applications due to the low parasitics of the transistor. Cellular and WIFI switches are now pervasively built in large lithographic SOI technologies. 45nm PDSOI has been widely investigated for many mmWave applications for phased array systems. The ability to stack transistors in PD SOI greatly increases the power handling and enables switches and power amplifiers to be built using low voltage CMOS devices. Fully depleted SOI extends the performance to even higher levels with a HighK-MG gate stack, 22nm gate length, and a thin silicon channel. This talk will describe the technology aspects that make SOI well suited for RF mmWave applications.

This presentation will briefly describe the structure and physics of partially depleted (PD), fully depleted (FD) and dynamically depleted (DD) SOI MOSFETs. The challenges and techniques of modelling each will then be discussed, with particular attention to the requirements for mmWave models. This will include all of the physical effects unique to SOI, including kink effect, self heating, non-quasistatic effects, body contacts, back gate modeling and substrate coupling. We will discuss the time constants associated with self heating and body effects. We describe the models for FEOL and BEOL line passives including lumped element passives such as varactors, resistors, capacitors, inductors and transformers as well as transmission lines components. We will also discuss the use of PCells, various choose for the PCell / PEX boundary and the incorporation of electromagnetic simulation into the design flow.

This presentation will discuss the main features of FD-SOI CMOS technology and how to efficiently use its unique features for RF and mm-wave SoCs. We will overview the impact of the back-gate bias on the measured I-V, transconductance, fT and fMAX characteristics and compare the MAG of FDSOI MOSFETs with those of planar bulk CMOS and SiGe BiCMOS transistors through measurements up to 325 GHz. Finally, we will provide examples of LNA, mixer, switches, and PA circuit topologies and layouts that make efficient use of the back-gate bias to overcome the limitations associated with the low breakdown voltage of sub 28nm CMOS technologies.

An overview of circuit design techniques and topologies for transceiver building blocks in deep sub-micron CMOS SOI will be presented. First, FET layout optimization considerations are given. Next, design examples of a 60GHz LNA, 24GHz VCO, 60GHz class-E PA, and a 60GHz t-line based phase shifter, all with state-of-the-art performance are provided. A fully-integrated 60 GHz transceiver in 32nm SOI CMOS, which integrates most of the above-mentioned building blocks, is presented to illustrate transceiver-level integration considerations. Finally, the challenges associated with process variability and test are outlined and examples of on-chip test and calibration techniques are given.

Has RF performance peaked; Are the glory days behind us? This panel will discuss the current state of silicon mmWave technologies and if we are really gaining much in RF with scaling across all technologies. What is the best silicon technology for RF/mmWave and where are we now with performance and scaling?

A wireless handset requires dozens of highly-selective narrow-band band-pass filters to couple signals that travel between the handset’s Antenna port and transceiver IC. These filters comprise a large portion of the front-end circuit of the handset and they are typically incorporated into a discrete RF front-end module along with power amplifiers, switches, and control logic. At the present time, SAW, BAW, and FBAR filters are the only known technologies with sufficiently low insertion loss and steep filter skirts which are available in sufficiently small hermetic packages for use as filter elements in front-end modules. The demand to incorporate additional filter functions increases with each successive handset generation. It has been suggested that tunable filters could be used to replace groups of fixed filters to either reduce the size of the front-end modules or to enable more filtering function in a module with a given size. This would also help to reduce the total cost of the front-end circuit. There are, however, numerous monumental technical challenges that make continuously and discretely tunable SAW, BAW, and FBAR filters impractical for use in front-end modules of wireless handsets. A common front-end configuration comprises a bank of filters which are switched in and out one at a time. A reconfigurable filter could be used in place of this switched bank of filters. However, for this to be feasible, the reconfigurable filter would need both frequency and bandwidth tuning “knobs”. To achieve this tunability, schemes have been proposed in which the frequency and kt2 of the constituent resonators of the filter are tuned in discrete steps. For example, banks of switched capacitors in series and in parallel with each resonator could be used for tuning; unfortunately, such banks of capacitors would significantly reduce the Q of the resonators, would occupy significant area and would reduce the power handling capability of a filter. Other schemes have been proposed in which the frequency of the constituent resonators is controlled by a dc voltage or magnetic field. Such devices have poor Q, don't have kt2 tunability, and are inherently non-linear, all of which are seriously performance limiting in the RF front-end. Another idea which has been proposed is to use an ultra-narrow tunable filter with steep filter skirts and a wide tuning range which could be tuned to a particular channel of interest. The problem with such a scheme is that modern 4G and LTE modulation protocols, require that the communication bandwidth vary from 0.2 MHz to more than 20 MHz and constructing a low insertion loss 20 MHz wide tunable filter with sharp filter skirts at GHz frequencies is unfeasible. The tunability problem is further exacerbated in the most recent handset generations where Carrier Aggregation (CA) functionality requires a handset to be capable of receiving multiple signals simultaneously. To enable CA functionality, a set of custom multiplexers is required. Each multiplexer: (1) is comprised of several filters which are attached to a single ANT port through a custom matching network and (2) must meet numerous RF specifications of the handset including not only the individual filter insertion loss, return loss, and isolation specs, but also a set of cross isolation specs between filters. Attempts to design a tunable version of a multiplexer comprising several tunable filters would be impractical because each tuned state of the multiplexer would need to comply with a different set of stringent spec requirements. Specifically, the fixed matching network will prevent the Tx and Rx ports of the multiplexer from having a tight "spot size" (return loss) over the full range of tuning, while the fixed cross couplings in the filter will prevent the precise frequency-placement of the poles and zeroes of filter function over the full range of tuning. Therefore, for the present and medium term, traditional filtering is the architecture of choice. As new bands and as new functionality such as 4-way MIMO are incorporated into RF front-ends, the number of filters per module will continue to increase. To respond to this pressure, technological innovation in piezoelectric filters must enable further reduction in filter size. A review of the evolution of this size scaling will be discussed as well as a discussion of module architectures which allow re-use of certain filters.

In current mobile communication equipment, a large number of filters and duplexers using surface and bulk acoustic wave (SAW/BAW) technologies are used to support multi-band and multi-standard operation. The number does not seem to be saturated. The increase made the RF frontend so complex, and introduction of MIMO and CA accelerate this trend further. One possible solution to overcome this problem is giving tunability to such devices. For ultimate down-sizing, use of MEMS technologies is mandatory not only for realization of tunable passive components but also for their hetero-integration with SAW/BAW devices and/or RF ICs. This talk discusses tunable SAW/BAW devices using the MEM technology. It is shown what are possible and what are not by using this approach.

For Software Definied Radios and other reconfigurale RF front-ends, flexibly programmable high-Q bandpass filtering and frequency conversion is a challenge. High linearity and blocker tolerance is crucial for interference robustness. Passive switch-R-C circuits, also known as N-path filters or Frequency Translated filters, can implement the desired functionality. They can both offer tunable filter functionality, but also frequency conversion with built-in RF-filtering. N-path filters essentially realize a low-pass R-C or high-pass C-R function in baseband, which is frequency translated to a band-pass or noth filter. As passive switch-R-C networks are used, linearity can be good. Moreover, the frequency translation is controlled by a digital clock, allowing for flexibly programmable very wide tuning range covering more than an octave or even decade in frequency. The resulting N-path filter benefits from CMOS scaling as switch parasitics improve, and increasingly higher digital clock frequencies are feasible. This workshop contribution will review the developments in CMOS N-path filters over the last decade, highlighthing promising achieved results, while also discussing performance limitations and challenges.

The talk will present the latest in RF MEMS tunable networks using the Cavendish RF MEMS varactors. Very high linearity and low loss tunable bandpass filters and notch filters in the 1-10 GHz region will be presented.

Adaptive and reconfigurable radios that can change their frequency and mode of operation based on the unused/available wireless spectrum as well as their surrounding environmental conditions have been proposed to address such challenges. However, currently available RF and microwave circuit components cannot meet the performance requirements, and cost constraints necessary for the commercialization of such systems. This presentation is on the applications of ferroelectric thin film barium strontium titanate (BST), a low loss, high dielectric constant field dependent multifunctional material. The electric field dependence of BST has been employed to design tunable RF and microwave devices and components. Another important characteristic of BST is its DC electric field induced piezoelectric and electrostrictive effect. These properties are utilized to design intrinsically switchable film bulk acoustic wave resonators (FBARs) and FBAR filters. Switchable ferroelectric based filter banks can significantly reduce size and power consumption of conventional filter banks employed in multi-standard and frequency agile radios. Properties and performance of several BST based adaptive and reconfigurable RF circuits will be presented.

4G networks integrators are demanding multiple RF tests which require large dynamic range and versatility. The continuous evolving of LTE bands are presenting a challenge to the design engineers, in terms of multiple modulation bandwidth and wide band span from 700MHz to 6000MHz. For Passive Components in the High Power environment, such as; Antennas, Filters, Cables Couplers, etc. the first and foremost requirement driving the dynamic range level is the PIM. That is, Passive intermodulation product due to two 20W carriers in the transmit band producing products in the Receive band, is required to be in the order of -165dBc. For components in the mobile side, such as: Switches, Filters, F.E.M, Etc. The IMD level is obviously higher, but other tests such as; Carrier Aggregation, Load-Pull, Harmonics, Jammers induced and more, are being conducted. Here, the tunable filter approach, often digitally controlled for flexibility, is considered. This presentation attempts to put together new block diagrams based on tunable and/or fixed filters to provide large dynamic range and versatility. In short, a cost effective system that can measure across multiple LTE bands and that can be expanded in a “Plug and Play” style.

While a plethora of center-frequency-tunable filters have been introduced over the last several years, transfer-function reconfiguration is significantly harder to achieve due to lack of appropriate architectures. New architectures are needed to address this need. It is the purpose of this talk to discuss such architectures that result in fully-reconfigurable bandpass and bandstop filters. Special attention will be paid to advanced topologies that enable reconfiguration of both poles and zeros of the filters. Besides single-band systems, multi-band reconfiguration techniques that allow independent control of each band will be reviewed. From a technology point of view, both planar and miniaturized cavity-based technologies will be discussed. Specifically, we will present proof-of-concept prototypes based on varactor-tuned microstrip resonators as well as tunable evanescent-mode cavity resonators. This talk will also briefly discuss the question of feedback versus open-loop control operation.

The first part of the talk is dedicated to reconfigurable Substrate-Integrated –Waveguide (SIW) filter and antenna structures and describes their tuning methodology and performance at microwave frequencies. The second part of the talk is dealing with design and implementation of tunable fluidic microwave filters and antennas, their advantages compared to other tuning techniques and their performance at microwave frequencies.

This talk will present the recent development on reconfigurable filter/antenna and antenna arrays. The first part of the talk will focus on designing the microwave filter and antenna as an inseparable unit which can achieve compact size and high efficiency. Tuning mechanisms are introduced in this filter/antenna so that the center frequency can be tuned. Special efforts are made to maintain the impedance matching across the tuning range by considering the frequency-dependent coupling coefficient. Both planar and 3-D designs will be presented. The second part of the talk will focus on reconfigurable antenna arrays which exhibit wide frequency range, continuous frequency coverage and large instantaneous bandwidth.

Phase change materials (PCM) are of great interest as they exhibit a phase transition from a semiconductor state to metal state through either thermal or optical excitation. They exhibit resistivity changes of several order of magnitude as they change state exhibiting a thermal reversible transition such as Vanadium Oxides (VO2) or a thermal switchable-latching transition such as Germanium Telluride (GeTe). While such materials have been widely employed in optical applications, Only very recently, there have been extensive research efforts to use them in RF applications. In particular, PCM-based RF switches combine the low insertion loss performance of MEMS technology and the small size and reliability performance of semiconductor technology. This talk addresses the use of both MEMS and PCM-based switches in the development of high-Q tunable filters.

One of the challenges of the future wireless networking systems is the delivery of the high throughout data connection to every cities, streets, and corner. While demand for data increases exponentially year after year, the distribution network to support such data demand is limited to the existing fiber and wireless connections. In this presentation we discuss how to provide high density connectivity as an extension to the fiber network utilizing mmWave modular antenna array systems and architecture. We discuss a flexible architecture that expands as the demand for network density increases.

Active Electronically Scanned Array (AESA) systems are becoming a critical component for 5G backhaul and infrastructure applications. This workshop describes the challenges that these applications present, how highly integrated silicon RFICs can meet those challenges and provide the requisite beam steering, amplitude taper, low noise amplification, and transmit power functions to enable planar arrays at 28 GHz and 39 GHz. Imbedded in the RFICs are an array of features that provide a high level of system flexibility not previously achievable in commercial AESAs.

Several radio concepts with the capability of providing a multi-gigabit transmission rate are currently investigated for 5G Access and Backhaul networks. Millimeter-wave communication is seen as one of the most promising key technologies, because of the wide spectrum available in this band. In order to be successful, High efficiency, high performance and low cost at mm-waves are required in the transceiver design, which poses several challenges for a MMIC design option. These technical challenges are discussed and some design solutions are presented which cover the frequency spectrum from Ka-Band to D-Band.

The constant increase of data traffic on the mobile communication network is pushing the manufacturers of digital radio links employed in wireless backhaul to evolve their products following some main guidelines: the adoption of more and more complex and spectrally efficient modulation schemes, the implementation of more sophisticated frequency reuse methods (i.e. MIMO in conjunction with XPIC), and the extension of the used frequency range beyond the e-band, where wider channels are available. Every step in these three directions is a challenge for both the microwave engineers and the DSP designers: the latter are constantly called to improve the digital algorithms employed to face not only the radio propagation issues, but also, more and more frequently, the effects of the analog hardware imperfections. This presentation talks briefly about the 'state of the art' of the modulation and coding schemes employed in digital radio links, and gives an overview of how the RF impairments are faced in the digital processing world while the complexity and performances of the radio equipment grow.

The backhaul networks are headed for a big change for coping with the massive increase of data that 5G brings. Higher bandwidths, more complex modulation formats and higher output power put challenging requirements on the hardware for achieving higher data rates. Transmitter and receiver architectures may look quite different from today’s line-up because of the tougher requirements on linearity, noise, efficiency and dynamic range. For the power amplifier in such equipment, the most critical parameters are output power and linearity to be able to use high modulation formats and higher output power for longer distance communications. Digital pre-distortion is nowadays common practice to linearize amplifiers at lower frequencies. Usually for microwave radios, this technique is implemented in the digital domain at the expense of increased bandwidth in the digital/analog converter (DAC). At mm-wave frequencies, when the signal channel bandwidth is much wider, the increased signal bandwidth used in the DAC will consume too high power, making this technique inappropriate to use for enhancing the radio performance. Analog pre-distortion of power amplifiers on the other hand can be implemented at mm-wave frequencies with little or no added power consumption to enhance the linearity. III-V technologies offer the best RF performance. GaAs High Electron Mobility Transistor (HEMT) is the most used and established technology for mm-wave MMICs due to its satisfactory high frequency performance. The reliability, stable manufacturing, performance and pricing of GaAs are very competitive, which make this technology very attractive for commercial use. High bandgap materials such as GaN are emerging technologies that offer significantly higher output power and linearity and is a candidate for being the next generation’s technology. We present the challenges on circuit design, system solutions and linearized PA design for the future frontends in the 5G backhaul networks using III-V technologies.

The forthcoming 5 generation of mobile will strongly affect the whole network infrastructure, including the backhaul. Microwave and millimeter wave radios will be widely employed in future backhaul deployment. The power amplifier represents a crucial components in these radios, and it will probably need to shift its paradigm in terms of required power, bands, frequency, to cope with the new scenario. This talk will shortly introduce the present situation and the foreseen trends regarding the possible evolution of this fundamental block of the infrastructure.

Next generation 5G backhaul infrastructures requires large uninterrupted bandwidth to support high capacity wireless communications. Both 60GHz and E-Band offer the required bandwidth under unlicensed/lightly-licensed regulation. Further enhancement of data throughput is achieved by increasing the modulation order limited by linearity and noise performance of the transceiver. We will review fully-integrated low cost fixed beam high-performance SiGe based transceiver chipsets for the full band between 57-66GHz, and for the upper and lower E-Band regions. Full duplex throughput is used to utilize the full potential of the band standards in a point-to-point configuration. Both designs were optimized for high modulation up to 256QAM and high power to support above 10Gbps while meeting outdoor regulation set by FCC and ETSI. In addition, a 60GHz phased array for PTP and point-to-multipoint applications will be presented. The phased array is in a TDD configuration with extremely high power / linearity performance and low noise.

A modular beamforming architecture utilizing multiple beamforming ICs and a separate IQ modem IC is a very efficient solution for scalable 5G backhaul networks. The beamforming IC consists of amplifiers and vector modulators, and the modem IC includes up- and down-converters with an integrated PLL. The proposed solution is highly compact, and can be expanded in a modular fashion, making it suitable for communication in small-cell backhaul networks. In this presentation, IC development in 130-nm SiGe BiCMOS process as well as IC-/system-level characterization will be discussed.

In digital power amplifiers the analog signal is encoded in a pulse train and restored only at the output by a bandpass filter. In theory, this approach allows for PAs with high efficiency also at backoff, which has yet to be demonstrated in practice. Overall performance depends on the combination of coding/modulation, the PA circuit plus filter, and the resulting linearity behavior, which all differ significantly from the classical analog case. The presentation reviews the state of the art of such PAs targeting the wireless infrastructure. The key factors determining transmitter performance are discussed and recent results on novel modulation schemes and GaN PA with improved PAE are presented.

A single-bit digital transmitter, in which the RF digital stream including the wireless signal is generated and fed to the antenna after amplification and filtering, is a promising candidate for the next generation mobile communication systems since it offers high flexibility required for multi-mode/band transmitters. In addition, it has the great advantage to significantly improve power efficiency of the transmitter when a switch-mode power amplifier (SMPA) is employed at the final stage. In this talk, we introduce highly-efficient and –linear single-bit digital transmitters using envelope delta-sigma modulation architecture without full DPD and digital Doherty transmitters which enhances back-off efficiency by digitally controlling two SMPAs combined with each other in H-bridge configuration. Finally the future applications of digital transmitters, such as digital radio-over-fiber system or software-defined radio with FPGA-based all-digital transmitter will be presented.

Multi-band multi-mode operation of 4G/5G broadband mobile communication creates several new design challenges to traditional RF transmitter architectures. For reconfigurable and flexible spectrum usage with compact and low-cost implementation, digital-intensive transmitter architectures have been recently proposed with high-efficiency power amplifiers in compound semiconductors. 4G/5G carrier aggregation, particularly non-contiguous multi-band transmission, demands staggering design requirements on bandwidth and linearity, which are difficult to manage with conventional digital modulation techniques such as pulse-width modulation (PWM) and delta-sigma modulation (DSM). In order to ovrecome these challenges of digital-intensive transmitters for non-contiguous multi-band transmission, we present advanced power encoding techniques, demonstrated with a proof-of-concept GaN HEMT Class-D digital outphasing RF power amplifiers. For broadband operation, the quantization noise of digital modulation often becomes performance bottleneck, consequently requiring high-order RF output filters with high insertion loss. We present novel out-of-band noise cancellation techniques for digital-intensive transmitters.

Massive MIMO systems are one of the key technologies for 5G. By using a large number of antennas, it is possible to increase the number of users with beamforming and optimize the channel capacity by using orthogonal signals. However, the cost of a transmitter with tens of antennas is usually very high and the implementations tend to be complex and bulky. Moreover, by using independent radios we have to deal with special synchronization hardware. In this talk we present some recent developments on all-digital transceivers that use several Multigabit Transceivers existent on medium-/high-performance FPGAs. We have built a proof of concept system using a single FPGA with 8 transmitting antennas, with a carrier frequency of 2.5GHz. The phase alignment of the antenna signals is performed on the baseband. This phase array antenna has a measured beam angle resolution below 1 degree. In this architecture, besides the FPGA we only need one filter for each channel, leading to a very compact and low-complexity solution when compared with other proposed systems. We also present an overview over several FPGA-based Delta Sigma Digital-to-Analog Converters with a larger bandwidth and higher flexibility that can be included into the same system. For the receiving part of the antenna array we also show some recent results on an all-digital 1-bit RF PWM analog to digital converter that is a good candidate to build a a complete All-Digital Antenna Array Transceiver Module that can be used for a diversity of different application scenarios: MIMO, Phased Array Systems, Radar Systems, etc.

The continuing demand on power, performance and cost for wireless handsets has spurred immense research and development for digitally-intensive transceivers. Yet, the ever increasing data rate and frequency bandwidth exacerbate major challenges to realize digital transmitters suitable for every use scenario. A variety of architectures have been proposed to overcome the hurdles. This presentation will first review the trend of digital transmitter architectures. In light of the current and emerging wireless standards, new opportunities and obstacles will be discussed. The works to date represent a beginning of ongoing development for future-generation all-digital transmitters.

Results from a project targeting mm-wave 5G transmitters are here presented. The focus in this workshop presentation is put on two 28GHz power amplifiers (PAs) and an RF-DAC, designed in an ultra-thin body and buried oxide (UTBB) fully depleted silicon on insulator (FD-SOI) 28nm CMOS technology process. The SOI technology enables stacked devices to be used in the PAs where back-gate biasing is utilized. Partially floating cascode gate performance on ACLR and EVM is also presented. Additionally, a 10-bit RF-DAC using a digital upsampling filter to achieve 1GHz RF bandwidth with more than 45dB SNDR is demonstrated.

The switched capacitor power amplifier (SCPA) or radio frequency (RF) capacitive digital-to-analog converter (C-DAC) combines the functionality of a mixer, a digital-to-analog converter, and a power amplifier (PA). The superior amplitude-to-amplitude (AM-AM) and amplitude-to-phase (AM-PM) linearity makes the SCPA attractive for implementations in mobile communication systems for latest communication standards such as Universal Mobile Telecommunications System (UMTS) or Long Term Evolution (LTE). In this talk the principle idea of the C-DAC as a configurable capacitive voltage divider, which additionally performs the mixing operation in dependency of the applied LO carrier signal, will be introduced. Two different transmitter structures, the IQ C-DAC based architecture and the polar C-DAC based architecture will be discussed. The C-DAC consists of an array of switched-capacitor cells, where each cell ideally consists of a CMOS inverter and a capacitor. An ideal C-DAC is perfectly linear. However, non-ideal components, i.e. switch parasitics and variations of the capacitors in the cells, cause the C-DAC to become non-linear, and thus generate AM-AM and AM-PM distortions. In addition, imperfect power supply sources generate unwanted harmonics in the C-DAC’s RF output signal. We present a switched non-linear state space model (SSM) which allows studying these effects with significantly reduced simulation run-time compared to circuit simulators. A comparison between the non-linear SSM, a transistor-level circuit model, and measurements on a 28 nm CMOS test chip is given for single tone as well as modulated LTE test signals. Furthermore, we will discuss digital pre-distortion concepts to improve the spectral regrowth behavior and the error vector magnitude (EVM) of an IQ C-DAC based architecture. The approaches have been validated using the non-linear SSM as well as circuit level simulations, and by measurements with the 28 nm CMOS test chip.

The central challenge of all-digital transmitters is the representation of the wanted RF signal by digital pulse patterns. This talk gives an overview over various encoding methods starting with analog time-continuous RF pulse-width-modulation up to a purely digital FPGA-based approach suited for Massive-MIMO systems. The FPGA-based system is capable of meeting the linearity requirements for modern mobile communication systems using individual multi-gigabit-transceiver ports. This is achieved thanks to an advanced combination of PWM and DSM preprocessing and a dedicated calibration algorithm. All presented concepts are evaluated based on simulations and measurements with focus on coding efficiency, signal quality and spectral emissions.

Future 5G standard at higher frequencies will ask for complex modulation scheme and antenna network to provide the requested high data rate. One of the solutions is to use beam forming to share the information over a huge number of elementary cells and propagate the total signal thanks to the antenna network. To do this, we need to ensure that all the front ends will perform in parallel. This represents a very interesting challenge. Indeed, the beam forming will be efficient if each of the elementary cells maintains its nominal “average” performance. This presentation will focus on a new approach of the PA design, to target this point. Particularly, we are focusing in a self-contained behavior of the PA, that allows it to tune and sense its performances regarding its environment, its own behavior, and the other PA ones. Even if this challenge will probably ask to reduce a little the state of the art performance, it would definitely warranty the final network one.

Despite their great potentials, the use of mm-wave devices has been limited by the precision required in their implementation. In this talk we discuss several examples of how the dynamic reconfigurability of such system can lead to improved performance, robustness, and added functionality. We will discuss several examples such as dynamic polarization control and self-healing circuits.

This presentation will explore fully digital architectures and circuit topologies for future wireless backhaul systems with aggregate data rates comparable to those of future 64Gbaud fiberoptic systems. Potential circuit topologies in 45nm SOI CMOS, 55nm and 28nm FDSOI SiGe BiCMOS will be reviewed along with measurements of digital transmitters with free space constellation formation at 100 GHz and 140 GHz. Predistortion and spectral shaping techniques in the transmitter, and receiver ADC-based equalization at 64 GBaud will also be discussed.

It is estimated that billions of silicon-based RF power amplifiers (PAs) are already in RF front end modules (FEMs) for 3G/4G handsets, WLAN, and other wireless applications today. The III-V semiconductor-based RF PAs, however, can still offer superior frequency and breakdown performance with higher output power (POUT) and power-added-efficiency (PAE) and faster time-to-market, but silicon-based RF PAs do have the advantages in offering higher monolithic integration with added functionalities (e.g., flexible on-chip digital control and selection on power level, modulation schemes, frequency bands, adaptive matching, predistortion, etc.), which can translate into lower cost and smaller sizes attractive for broadband multi-mode multi-band handset transmitters. The advancement from 4G into 5G will for sure increase the complexity for PA design, as the higher RF signal modulation bandwidth (e.g., 250/500MHz and above 1 GHz) for transmitters, stringent linearity and efficiency requirements at cm-Wave and mm-Wave carrier frequencies (e.g., 15/28/38/45/60/73 GHz), multiple antennas for beamforming and massive MIMO, etc. will be particularly challenging for both 5G PA design and testing. The peak the multiple carriers and clusters may result in more challenging waveforms with high peak-to-average-power-ratio (PAPR) for transmitters. The greater numbers of the supported bands, MIMO antennas and 5G small cells will require many more RF PAs and make the high PAE performance and low-cost particularly attractive; and the significantly reduced 5G PA peak POUT requirements appear to be favorable for silicon-based PAs implementation. Therefore, some potentially key design techniques for high-efficiency 5G broadband wireless PA design will be discussed in this talk. The design and testing challenges ahead for highly efficient linear wideband RF/mm-Wave PA are particularly serious for 5G applications, especially for mm-Wave and massive MIMO-like scenarios. I will try to address in this talk why some these very difficult challenges can, with or without the high-performance GaN/GaAs PA, also present as golden opportunities and great incentives for silicon PA development and advancement in market segments.

Achieving a very high bit rate up to multi tens of gigabits per second is one of the goals of fifth generation (5G) mobile communication networks. Higher frequency bands such as millimeter-wave band are identified as a promising avenue because they will provide a contiguous broadband spectrum. To utilize the higher frequency bands, phased array antenna systems with high antenna gain and beam-forming capability will be a key technology. This presentation reviews several schemes to control antenna beam direction on a phased array antenna system for 5G utilization, and introduces the recent research results at NTT DOCOMO laboratories.

A moving field inductive power transfer (MFIPT) system for supplying power to electric vehicles while driving along the route is described. This MFIPT system uses primary coils arranged below the pavement. The primary coils transmit the energy via an alternating magnetic field to a secondary coil located at the vehicle below its floor. Only those primary coils located below the secondary coil of a vehicle are excited. By this way losses and radiation in the environment are minimized. The operation principle of the moving field inductive power transfer system is based on a switched DC-to-DC converter which converts the DC power supplied by the stationary power line to DC power delivered to the moving electric vehicle. The dynamics, the operating regimes and the power balance of the moving field inductive power transfer system and the costs for the implementation of the system are discussed. The contactless power supply of electric vehicles on highways makes it possible to get along with battery capacities otherwise suitable only for shorter range. The batteries are used only in local traffic and on side roads where no moving field inductive power transfer system is installed. In areas where there are no inductive supply roads available, the inductive energy transmission system may still be used in stationary charging stations. Since only the primary coils below the vehicles are activated, high efficiency is achieved and the magnetic field is shielded against the environment. The MFIPT system is especially interesting for intelligent autonomous electric vehicles. In transportation systems based on this combination the electric vehicles will exchange information with traffic management systems and with each other and thereby achieve a steady, energy-efficient traffic flow even at very high vehicle densities.

The ability to provide power without wires was imagined over a century ago, but assumed commercially impractical and impossible to realise. However for more than two decades the University of Auckland has been at the forefront of developing and commercialising this technology alongside its industrial partners. This research has proven that significant wireless power can be transferred over relatively large air-gaps efficiently and robustly. Early solutions were applied in industrial applications to power moving vehicles in clean room systems, roadway lighting, industrial plants, and in theme parks, but more recently this research has helped develop technology that has the ability to impact us directly at home. The seminar will describe some of the early motivations behind this research, and introduce some of the solutions which have been developed by the team of researchers at Auckland over two decades, many of which have found their way into the market. It will also describe how the technology has recently been re-developed and is evolving to enable battery charging of electric vehicles without the need to plug in, and alongside this how it has the potential to change the way we drive in the future.

Wireless power transfer (WPT) is one of the most promising technologies in recent years, and the transportation systems with WPT technology are expected to create huge market in near future. As the transportation systems require high power over kilo-watt, the design should be differentiated from low power WPT systems. As the magnetic field strength is much larger than any other electronic system, the electromagnetic safety issue becomes crucial especially in railway WPT system where mega-watt of power is required. In this talk, wireless power transfer systems in vehicular applications and recent researches on electromagnetic safety for these high power WPT systems are introduced. Transmitting and receiving coil design, magnetic field shaping, and resonant reactive shield to minimize the leakage magnetic field in high-power low-frequency magnetic resonant WPT system are explained and related issues on standardization and commercialization are discussed.

Maximum link efficiency for mid range IPT systems often requires the system to operate at several MHz. This means significant challenges in the design of the electronics that drives the transmit coil and for the rectifier on the receiving coil. Achieving high power handling capability and efficiency at MHz frequencies is difficult with silicon devices and so this is an application where GaN and SiC transistors are especially well suited. A further requirement for high efficiency is the need to for the circuits to be soft switched and to be tolerant to changes in the geometry of the magnetic link. In this talk I will describe soft switched topologies suitable for high efficiency across a range of magnetic links, that reduce the stresses on the components and have minimal requirements for closed loop control bandwidth. The circuits are particularly suitable for ensuring ICNIRP regulations are met, even as the magnetic link geometry changes.

Inductive wireless power transfer (WPT) systems have traditionally been used for high-power large-gap near-field applications, such as electric vehicle (EV) charging. However, capacitive WPT systems can potentially be more efficient and less expensive as they do not require ferrite materials for flux guidance. This presentation addresses the design challenges and the tradeoffs associated with the design of a multi-modular capacitive WPT system suitable for stationary and in-motion EV charging. The system utilizes relative phasing of the different modules to achieve near-field field-focusing and hence maintains fringe fields within safety limits. The WPT system also requires matching networks that provide large voltage or current gain and reactive compensation. An analytical optimization approach for the design of L-section multistage matching networks is developed and utilized to maximize the matching network efficiency. The results of the proposed approach are validated using a 12-cm air-gap 6.78-MHz capacitive WPT system.

A nonconventional exploitation of a self-resonant near-field link at UHF for data communication, is combined in a compact inductive wireless power transfer system. At LF, the inductive channel is designed to deliver up to 1.3 kW to a resistive rotary heater. At UHF, sensing capabilities are made possible by exploiting self-resonant structures, such as split-ring resonators, one at each far-end side of the link. This network is used in a passive sensing system, to convert the data of a remote temperature sensor, representing the system variable load. The reflected power variations at the transmitter side, due to the dc load variations, are successfully used to perform the sensor readout.