Dr Duixian Liu, Yorktown Hights, US is a researcher at IBM at Thomas J. Watson Research Center since April 1996. His research interests are antenna design, electromagnetic modeling, digital signal processing, and communications technology. He received the IBM’s outstanding technical achievement awards in 2001 and 2002, and the IBM’s highest technical award, in 2003, for contributions to the integrated antenna subsystems for laptop computers.

Dr Ullrich Pfeiffer, Siegen, Germany is the head of the THz imaging group at the Institute of High-Frequency and Quantum Electronics at the University of Siegen, Germany. From 2001 to 2006 he was with the IBM T.J. Watson Research Center where his research involved RF circuit design, power amplifier design at 60 GHz and 77 GHz, high-frequency modeling and packaging for millimeter-wave communication systems. He is a member of the German Physical Society (DPG), and was the recipient of the 2004 and 2006 Lewis Winner Award for Outstanding Paper at the IEEE International Solid-State Circuit Conference. He received the European Young Investigator Award in 2006.

Dr Brian Gaucher, Yorktown Hights, US is a research staff member at the IBM T.J. Watson Research Center where he manages a communication system design and characterization group. His present research interests include 60 GHz Gbps wireless communication design and biomedical applications of wireless technology. His group has helped more than five products come to market. He is an IBM master inventor and holds two outstanding technical achievement awards and one corporate award.

Dr Janusz Grzyb, Pfaeffikon, Switzerland works at Huber + Suhner AG, Switzerland since 2006 as a senior R&D microwave engineer responsible for the development of a series of 60 GHz products. Before then, he worked at IBM T. J. Watson Research Center, NY. His primary responsibilities there were antenna and package design for 60-GHz wireless systems

Advanced Millimeter-wave Technologies

Antennas, Packaging and CircuitsBy Duixian Liu Ulrich Pfeiffer Janusz Grzyb Brian Gaucher

John Wiley & Sons

Copyright © 2009 John Wiley & Sons, Ltd
All right reserved.
ISBN: 978-0-470-99617-1

Chapter One

Introduction

Brian Gaucher

There is an unusual confluence of three major disruptive and threshold events taking place that are fundamentally reshaping the wireless industry. First, wireless has become a critical, accepted and necessary part of everyday life, e.g. the number of mobile phone users growing at approximately one billion per year and younger generations and countries skipping the PC in favor of handheld wireless devices. The second threshold event is the now rapidly growing high-definition video, automotive radar and high-resolution imaging markets, which have created a sudden need for extremely broadband gigabits per second (Gbps), highly integrated, low-cost and low-power wireless devices in the millimeter-wave (mmWave) frequency bands, which previously only the military could afford. The third threshold event occurring is silicon technology and tools have been developed with suitable performance characteristics to enable radio design, integration and operation at mmWave frequencies; here we specifically discuss the range 60-194 GHz, although the literature is now showing silicon working at hundreds of gigahertz.

The enormous reliance of consumers and enterprises on wireless and the evolving mobile Internet is having a disruptive influence on the telco industry, e.g. wireless subscriptions now outnumber landline phone subscriptions and Apple’s iPhone(tm) or Google’s GPhone(tm) are forcing mobile carriers into ‘openness’, effectively reinventing wireless networks and the way they operate and make money. This reliance and evolution speaks to the convenience factor as well as perceived new utility that this technology is providing consumers and enterprise users.

Likewise in the automotive industry, radar units are now options for high-end vehicles and may become mandatory under Intelligent Highway programs around the world, owing to the potential increased safety they can provide. This means a new market for tens of millions of mmWave systems per year.

The growing high-definition multimedia revolution will also require a significant portion of devices to be wireless. This is creating demand for high-speed bandwidth that goes well beyond what wireless systems of today can handle. If one looks at the lowest bandwidth requirement for uncompressed high definition television (HDTV), it is about 1.5 Gbps and with some minor coding to make it more robust to multi-path and fading, it easily tops 2 Gbps as discussed within the IEEE 802.15.3c. Today’s conventional WiFi delivers unprecedented performance for both office and home use, but tops out at 54 Mbps, with some proprietary systems going as high as 108 Mbps. There is hope that 802.11n and ultra wide band (UWB) systems will get as high as 480 Mbps. UWB systems that utilize limited spectrum between 3.1 and 10.6 GHz have met with marginal success so far, but also promise up to 960 Mbps over a meter or two. At the time of writing, such systems have met with only marginal success in both data rate and distance. Thus, wireless is exploding in use and rapidly evolving from convenience to need across multiple industries and will continue to grow in this direction. In order to satisfy these future WiFi, HDTV, radar and other system needs of speed, capacity, security and robust performance over distance, completely new mmWave (60-194 GHz) solutions will be required.

1.1 Challenges

There is a growing need to solve the huge technical hurdles to cost-effectively leverage the vast unlicensed bandwidth available at mmWave frequencies and satisfy these growing demands. Wireless HDTV is a good example of this and worthy of further exploration here. This application is demanding dramatically higher data rates on the order of 10-100 times current rates. Indeed, they are increasing much faster than current wireless systems can handle and an alternative solution is needed. Given the data rate, capacity and quality of service (QoS) requirements, this can only reasonably happen in a spectrum location where there is suitable worldwide bandwidth on the order of gigahertz with rules that allow one to close a reasonable link budget. These issues have been the fuel and motivation for looking upward in spectrum. As one climbs the spectrum ladder, the first frequency allocation where all of this has the possibility of working well across the varied application space is the 60 GHz band. At 60 GHz there is 3-7 GHz of worldwide bandwidth available depending upon the country; see Figure 1.1 for a sample of countries. In terms of available bandwidth and allowable rules such as transmit power, lack of incumbent users, simple flexible transmission rules, etc. 60 GHz is a boon but it also comes with significant challenges, e.g. ability to do this in a low-cost, physically suitable and robust manner; likewise, the 77 GHz band is the direction the automotive industry is taking, especially in Europe, for automotive radar systems, also called adaptive cruise control (ACC) and Collision Avoidance Systems. There is also increased development activity in the 94, 120 and 194 GHz bands which are being utilized for homeland security applications such as radar, imaging systems, remote sensing, active denial and many others.

Given this set of events, opportunities and constraints, wireless designers have begun developing mmWave system architectures, circuits, antennas and packages; but as expected, they face enormous challenges of simulation, design, integration, physical realization, packaging and test of complete systems that are literally orders of magnitude more difficult than 2.4 and 5 GHz WiFi systems of today; yet to be successful they have to be nearly the same cost.

If designers can find ways around these challenges, then there are significant benefits that are well worth the effort. The bandwidth and flexible open rules are the most obvious, but there are other less obvious ones also, depending upon if the glass is viewed as half full or half empty. At 60 GHz the wavelength in free space is approximately 5 mm, so circuit designers have the option to use transmission line structures as matching elements and resonant structures, in ways impossible to think about at 2.4 or 5 GHz. Similarly, on-chip filtering becomes possible and on-chip or in-package antennas are now a choice. Traditional microwave board elements such as Lange couplers, 90 hybrids, rat-race structures and many others are now small enough for on-chip consideration. Antenna beam-forming, steering and spatial-power combining become viable system design considerations even in consumer-level solutions, something previously only enjoyed by high-end military systems. Owing to this wavelength consideration, levels of integration go beyond what is achievable at 2.4 and 5 GHz, e.g. including filters and antennas on chip or in the chip package. Indeed, the whole area of packaging is flipped on its head when considering including antennas within the package itself. One’s packaging mindset shifts from containing radiofrequency (RF) radiation to intentionally radiating specific frequencies, while attenuating others. This will be one of the themes for this book.

Then there is the glass half empty point of view. At mmWave frequencies, the world of consumer level systems, circuit, antenna and packaging design, is largely unknown. For example, at 2.4 and 5 GHz, a designer takes the dielectric constant of PC boards for granted, something that is known and can be relied upon; but this is not the case at e.g. 60 GHz and higher. At such frequency extremes, each material has to be characterized and relevant data extracted from samples. There are also new metamaterials and devices that might prove invaluable, e.g. electronic band gap (EBG), anisotropic approaches, new polarization techniques, etc., but as yet they remain largely untried and untested at mmWave frequencies. Simple interconnects that work well at 5 GHz may have untenable loss at 60 GHz, where each and every interface outside the chip has to be considered in the link budget. As designers pull the antenna design into the chip or chip package, there are many new and important considerations and design options that need to be taken into account and traded off, which affect the whole system. Then there are the circuit design challenges too. At 60 GHz basic transistors, as good as they have become recently, run out of gas in this frequency range, e.g. gain is considerably lower than 5 GHz, isolation decreases, power generation is much more challenging and losses are much higher. On top of all of this, today’s design tools have not been tested in any significant way at 60 GHz and even the smallest variation or error may have a significant impact on the end product’s performance. Although time will help, today even relatively simple 60 GHz circuits require hours to simulate and it is not unusual to wait days for more complex circuits. Simulation times of digital systems that include analog front ends running at 60 GHz are currently un-simulateable due to both convergence and time required. And where do designers get valid active and passive models and macros they can trust for use in silicon level chip design that yield the correct designs at the first attempt? Even basic RF switches and switch elements need special consideration for device selection, design, use and simulation.

Thus, the challenges and benefits are many, making this an incredibly rich and deep area of research in the coming years, with streams of patents emerging around all of the above, virtually reinventing the wireless industry. As all of this takes shape, we will begin to benefit as consumers of this technology; we will find 60 GHz will enable 1-10 Gbps wireless solutions in a wide range of products such as next-generation WiFi, wireless HDTV, MP3 players and cell phones. These latter devices will be used as electronic wallets that can nearly instantly download, pay for and store a HD-movie for transport to the home and wirelessly upload it to home theater systems. There will be kiosks that act as displays for the hard drives and systems that reside in your cell phone and so much more.

As we master the design idiosyncrasies of 60 GHz, the wireless industry will move up to 77 GHz to develop cost-effective ACC for cars and Intelligent Highway Systems, where there exist a whole new set of packaging and antenna challenges on top of extreme environmental conditions. As next-generation silicon-based terahertz (THz) imagers evolve, inspectors will more easily be able to detect non-metallic weapons and explosives from a safe distance using THz imaging techniques and high-performance computing. This will make airports and critical entry points safer for everyone. However, to achieve all or any of the above in any cost-effective manner requires designers to solve a myriad of extremely challenging problems.

1.2 Discussion Framework

With a reasonable motivation of why the mmWave frequency is important and useful and the basics of what the challenges are, the next step is to establish a simple consistent framework that can be used throughout the book that addresses the system architecture, antennas, circuits and packaging in a holistic manner. Rather than highlight the multitude of architectural solutions for all wireless architectures and applications, e.g. direct conversion, low-intermediate frequency (IF), heterodyne, super-heterodyne, etc., and performance targets, we choose a single commonly applied approach of the super-heterodyne architecture. We use this as our reference point for discussion purposes. It is not the simplest architecture, but is general purpose and contains all the elements of virtually any wireless system one can imagine. Based on that architectural approach, we call out circuits, antennas and packaging options that reference this and which then be addressed in the remainder of the book.

1.3 Circuits

The super-heterodyne architecture is one of the more complex and therefore more encompassing architectures, making it a good choice to frame the larger circuit, antenna and packaging challenges. Although not optimal for all, it could be used for any of the aforementioned applications; Gbps wireless, radar, imaging etc. A simplified block diagram of a 60 GHz radio architecture is shown in Figure 1.2. It is based on a single-voltage controlled oscillator (VCO) super-heterodyne (multiple mixing stages) time-division duplex (switches from transmit to receive (T/R) rather than simultaneous transmit and receive) design with variable-IF frequency (byproduct of a single local oscillator (LO)). The high-frequency 60 GHz signals connect directly from integrated antennas to the T/R switch and to the low noise amplifier (LNA) input or the power amplifier (PA) output to avoid the need for external packaging and waveguide structures with their associated size, weight and power losses.

1.4 Antenna

For reasons to be discussed in detail throughout the book, mmWave antennas will nearly always be integrated with the chip or chip package. The integrated antenna has two major functions: the first is an efficient radiator or collector and the second as a bandpass filtering function. The natural bandpass filtering provided by the antenna helps both to limit the noise bandwidth prior to the LNA and to provide some image rejection. It is critical to any wireless system, but at mmWave frequencies the antenna becomes even more so because of the potential detrimental effects of interconnection losses, distance to RF electronics, match and proximity effects of nearby materials, receiver noise figure and transmit power. At mmWave frequencies power generation is extremely difficult and ‘expensive’ DC power wise, making antenna efficiency paramount. Designers cannot afford to waste hard-fought-for RF power only to lose it to ‘simple’ but lossy interconnects. For consumer level (cost-sensitive) products, performance needs and ease of use, it is a virtual requirement that the antenna be an integral part of the chip package from start to finish. This places new constraints across the whole antenna/package simulation, design and test space.

1.5 RF Electronics

The RF electronics make up the body of any wireless device and consist of a receiver and transmitter where each place critical and sometimes conflicting requirements on both the antenna and the package. For example, loss between the antenna and the RF electronics decreases transmit output power and increases system noise figure directly; thus, each RF subsystem needs to be located close to the antenna. Since there are physical limitations and both cannot be minimized, tradeoffs must be made as to which should be the closest. To establish a common vocabulary for the rest of the book, it is worth reviewing both receive and transmit chains.

1.5.1 Receiver

As shown in Figure 1.2, the signal at the output of the receive antenna is amplified by an LNA with enough gain (>10-15 dB) to establish the system noise figure (<6 dB). The LNA drives an integrated bandpass image filter, which is designed to eliminate received noise power amplified by the LNA at the IF image frequency to minimize the noise. The output of the LNA drives a mixer that translates the mmWave signal to an IF frequency in the range of 9.2 GHz. This IF frequency is chosen to provide easy image rejection at the RF and to support very high data rates. The mixer then drives a variable-gain IF amplifier, which increases the dynamic range of the receiver.

A single VCO operating in the band around 17.6 GHz is multiplied by three to generate the LO for the up and down mixers. The VCO is also divided by two to generate the 8.8 GHz LO signals for the IF quadrature mixers, which translate the received signal to baseband frequency I and Q channels. The baseband I and Q channels are band limited by variable bandwidth low-pass filters. From here the signals are either A/D (Analog to Digital) converted or detected directly depending upon the chosen modulation. The only cost effective, power efficient and performance friendly design approach at mmWave frequencies is to implement fully integrated solutions in silicon; going off chip for any mmWave frequency element could prove disastrous.

(Continues…)

Excerpted from Advanced Millimeter-wave Technologiesby Duixian Liu Ulrich Pfeiffer Janusz Grzyb Brian Gaucher Copyright © 2009 by John Wiley & Sons, Ltd. Excerpted by permission.
All rights reserved. No part of this excerpt may be reproduced or reprinted without permission in writing from the publisher.
Excerpts are provided by Dial-A-Book Inc. solely for the personal use of visitors to this web site.