EXECUTIVE SUMMARY

Anthony Ephremides and the Members of the WTEC Panel

INTRODUCTION

What follows is a summary discussion of the current state and trends in wireless communications. The overview of this rapidly expanding and technologically diverse field is based on the discoveries of the WTEC panelists during their site visits to leading companies and experts in Europe, Japan, and the United States as well as the panelists' expertise in the various areas of wireless communications. Included in this chapter are sections on the major drivers affecting wireless communications growth; a discussion of bottlenecks, such as systems interoperability, that threaten the development of wireless connectivity; and, finally, a sampling of technology highpoints, including the much-discussed "software radios."

The findings related in this study were made by the WTEC panelists listed in Appendix A. Table ES.1 (p. viii) contains a ratings chart that shows a general comparison of the three regions in the various technologies essential to the field, summarizing this panel's findings.

U.S. COMPETITIVENESS

On the question of U.S. competitiveness, which of course ranks high in the interests of the sponsors, the incomplete nature of the sampling, and U.S. and foreign reluctance to disclose fully, made a comparative assessment of capabilities rather difficult. In addition, the corporate world of communications (wireless or not) is especially globalized and international. It is increasingly difficult to identify national boundaries in the activities of various companies. As an example, consider the case of Nokia, probably the most successful manufacturer of cell phones at present, and one of the major players in the arena of wireless communications. Most of Nokia's long-term research is conducted in Texas, and the overwhelming majority of its stockholders are U.S. citizens. Yet, the corporate decisions are made in Helsinki.

In any event, the panel made a genuine effort to assess the comparative strengths and weaknesses in the production of technologies that are crucial to wireless communications in the United States, in Europe, and in Japan. The reader will find in several of the chapters that follow an evaluation based on numerical scores that attempt to characterize the relative position of U.S. industry vis-a-vis that of European and Japanese industry. Although the panel is making these disclaimers, it was able, in several cases, to provide such an assessment. Specific assessments that the panel made are summarized in Table ES.1 (p. viii). And, indeed, within each technical area there is some (but not much) variability. However, overall the state of development of wireless communication technology in all three regions is very comparable. Thus, it is safe to state that Europe, Japan, and the United States are engaged in a neck and neck race with increasingly blurred boundaries among their efforts, goals, and achievements.

In the early development of the second generation cellular systems for cellular voice systems (which is only one of the multiple facets of wireless), the European Global System for Mobile Communications (GSM) system was an early winner, while Japan was a distant third in the race. Of course, the converging third generation systems have brought all three regions close to each other. And, as stated earlier, there are many other aspects of wireless communications in which the regions and industry compete.

A remarkable observation, however, is that, in a general sense, there is a difference in attitude toward wireless technology development between the United States and Europe, on the one hand, and Japan, on the other. The panel found that in the "Western" world the attitude is preponderantly market-driven, while in Japan it seemed more technology-driven. As a result the panel found an increased readiness and imagination in Japan to use the technology to develop applications and diverse usages, while in Europe there is a more cautious and conservative attitude. In the United States there is a "happy medium" in that the commercial world shares the caution of the Europeans, while the military establishment is pushing the frontiers to apply every available bit of technology and to develop more technology.

Table ES. 1
A Comparison of Wireless Technologies and Networks in Europe, Japan and the United States

Technology

Europe Status

Japan Status

U.S. Status

Hardware

- mm wave circuits and systems

****

*****

**

- Packaging/interconnect

***

****

*****

- CAD

***

**

*****

- SiGE/Si

***** (Ger.)

***

****

- III-V

****

*****

*****

- GaN

***

****

****

- Antennas

***

****

***

- Passive components

****

*****

****

- Amplifier technique

**** (U.K.)

****

*****

- MEMS

**

***

****

Channel Characterization and Propagation Models

- Statistical/empirical

***

***

***

- EM based deterministic

**

*

***

- Integrative models

**

-

*

- Microwave and millimeter wave

*

**

**

Coding, Modulation and Multiple Access

**

***

*****

- Multi-user detection theory

**

***

*****

- Multi-user detection implementation

***

*****

***

- Coding

*****

***

*****

- Multiple access

****

****

****

Software Radios

- Software radio technologies

***

***

***

- Energy efficient communications

***

*

***

- Integrated approaches

*

***

*




MAJOR DRIVERS OF THE GROWTH OF WIRELESS COMMUNICATIONS

Major factors in the growth of wireless communications include the following:

MAJOR BOTTLENECKS IN THE DEVELOPMENT OF WIRELESS CONNECTIVITY

The Medium of Wireless Communication

There are two ways in which the medium raises serious technical challenges. The first one is the shortage of usable bandwidth. The expanding need for high-rate multimedia applications squeezes the available bandwidth in the frequencies at which propagation properties are favorable. Already most of the spectrum up to the low GHz range is used up. Beyond that, i.e., at Ka-band and beyond, optical phenomena, such as aerosol-induced dispersion, take hold and make communication at long ranges problematic. Yet, if the ultimate goals of wireless connectivity are to be realized, the limitations at the higher bands (up to 100 GHz) must be overcome.

The second limitation of the medium has to do with propagation and interference impairments. Multipath propagation (and the resulting inter-symbol interference), Doppler, shadowing, variable dielectric properties, etc. affect seriously signal strength and decoding ability. In addition, in the presence of multi-user communication, interference from other users represents a major obstacle. Of course, already there are technology directions that show promise toward overcoming these impairments (such as smart antennas, elaborate MAC protocols, and multi-user detectors); however, much more is needed.

Portability of Equipment

For the ubiquitous use of wireless communication it is necessary that the user equipment (currently referred to as the "handset" or the "palmtop") be extremely light, small, versatile, and above all, energy efficient.

Battery technology is only one of the means by which the longevity of portable devices can be assured. Pulsed use of the batteries, appropriate circuitry and packaging, directional antennas, suitable modulation and coding, and, most unexpectedly, intelligent network protocols can yield significant energy savings. Furthermore, system integration and space-sensitive design of signal processing algorithms can reduce the size (and weight) of the portable devices.

System Interoperability

One of the most disabling features in wireless communications is the simultaneous development of incompatible standards. Even in the widespread application of cellular telephony, the current systems in the United States, Europe, Japan, and the rest of the world are simply not compatible. Furthermore, if connectivity to the Internet is to be realized, the interoperability of diverse systems, such as satellite networks, wireless access based systems, the public switched telephone network (PSTN), etc. must be assured. Already progress has been made in the eagerly anticipated, so-called third-generation cellular systems toward the celebrated goal of convergence. However, interoperability must extend beyond that.

THIRD GENERATION CELLULAR SYSTEMS

The panel found that the major current preoccupations of most companies involved "third-generation" cellular systems. These systems defined most of the immediate concerns and shaped many future ones. Put simply, the first analog voice cellular systems (exemplified by the U.S. AMPS system) would be the first generation. The digital voice systems that followed (mainly the GSM in Europe and IS-136 and IS-95 in the United States) would constitute the second generation while the term "third generation" would refer to digital systems that superpose substantial data services. Of course the path to the third generation has been evolutionary, and "fractional" systems between the second and third generations can be identified.

The long gestation period that has occurred between generations can be attributed in part to genuine technical protocols, but primarily to economic and political difficulties that have to do with market share, intellectual property disputes, and other competitive aspects of product development and service provision.

It does appear certain that within the next couple of years a quantum step in the advancement of wireless communications will have occurred with the launching of third generation systems. But, then the question is "what next?" The panel did ask that question and obtained a variety of answers. Some view as "fourth generation" those systems that will be capable of delivering multimedia services at rates upward of 10 Mbps to mobile users. Such systems will probably need to operate at the higher GHz regions of the spectrum where there is ample bandwidth. Others are more skeptical, considering the whole issue of future generation systems as obscure. However, many of these skeptics are involved in funded studies on the very issue of fourth generation systems. Thus, undoubtedly, the next generation of wireless systems ranks high in everyone's mind. It was interesting that some (especially in Japan) ventured a guess about what the fifth generation wireless systems will look like. They presented a vision of unlimited bandwidth and of rates exceeding 1 Gbps at the 60-90 GHz range for the ultimate realization of global wireless connectivity.

WIRELESS ARCHITECTURES

Many people wonder what wireless architectures, other than cellular and wireless LANs, are possible or achievable in wireless networks. It has been known that at least in the United States, the military has shown a keen interest in the so-called "ad-hoc" network architectures. The panel found that although Europe and the United States showed rather limited interest in ad-hoc networks for commercial applications, there was more interest in Japan. Some imaginative applications for the deployment of civilian ad-hoc networks that involve large gatherings, stadium events, etc. were presented. A clear interest in ad-hoc networks of limited scope for home entertainment was expressed everywhere. The first extremely successful such architecture, which is intended for cable replacement at the office or at home, is the "Bluetooth" standard that has experienced overwhelming acceptance in industry worldwide. However, "Bluetooth" is a modest ad-hoc system. It involves only single-hop access to a "master" station and permits the participation of only up to seven "slave" stations. Especially in home entertainment networks, relaying and multi-hop operation will be required, since transmission power must be limited (because of a variety of regulatory and practical reasons) and since room and floor partitions induce significant attenuation.

FINDINGS BY TECHNICAL AREA

Within each of the technical areas of the study, the panel reached specific conclusions about the issues that researchers here and abroad agree require increased attention. These are reported in detail in the chapters of the main body of the report. Here is a brief summary.

Hardware

In the hardware area it is necessary to improve materials, components, and devices. In particular new materials such as GaN and SiGe need to be exploited further. In addition, microelectromechanical systems (MEMS) technology can offer advantages in building components. Extensions of the linear region for active devices as well as improved performance of passive devices are necessary. It is also necessary to develop broadband antennas of higher gain and lower cost phased arrays. In addition, exploitation of 3-D-oriented packaging technologies offers significant promise.

As far as front-end architecture is concerned, the tradeoff between amplifier linearization and efficiency must be optimized. Promise also appears in the development of multifunction/reconfigurable devices and circuits (as well as antennas) in order to facilitate the operation of so-called software-defined radio (a notion that has emerged as very prominent for the future of the field). Effort is needed to improve interconnects and packaging, while use of MEMS technology for radio frequency (RF) components promises significant gains.

Expansion of operation in the higher ends of the spectrum (10, 35, 60, 77, and 95 GHz) requires significant breakthroughs in all aspects of hardware mentioned above. The use of computer-aided design (CAD) tools for all-inclusive and interactive packaging that encompasses circuits, devices, and antennas, as well as electromagnetic, thermal, and mechanical functions, is seen as necessary.

Smart Antennas

In the area of smart antennas, the major improvements needed to realize better coverage and higher data rates are reduced cost and size (which requires improvements in both electronics and power consumption), increased use of diversity, and better tracking of mobile users. Some of the research directions that hold promise toward the achievement of these goals are the use of spatial-temporal signal processing algorithms, better and more versatile definitions of standards, and, above all, interdisciplinary, vertically integrated designs that incorporate hardware aspects, coding/modulation, access control, antenna directivity, switching, connectivity control, multimedia techniques (such as adaptive compression), and overall systems architecture. In fact, vertical integration emerged as another dominant notion for future wireless communication systems.

Channel Characterization

In the area of channel characterization, it was confirmed that impairments on system performance inflicted by the wireless medium continue to be a major bottleneck in the further development of wireless communications. At a minimum, it is necessary to understand, measure, model, and characterize the medium accurately before developing effective remedies.

As such, the key direction in channel modeling that the panel identified was the coupling of statistical with computational (empirical) models. For analysts and designers, the embodiment of channel characteristics into a few statistical parameters is very desirable and convenient as it simplifies the enormous task of overall system design and evaluation. Such models usually do not provide sufficient accuracy, but, on the positive side, they are not too dependent on specific environments.

On the other hand, computational models (such as ray tracing) are much more accurate but are very dependent on specific environments and require prohibitive amounts of computation. The only avenue for better channel characterizations requires imaginative coupling of these two methodologies. During its study the WTEC panel did identify vestiges of such coupling already.

Link Layer

In the area of link-layer operation and design, which includes coding, modulation, and access control, there are numerous ways in which wireless systems can benefit. After all, this is the most mature area in the "systems" aspects of wireless communication, since it has been the cornerstone of classical point-to-point transmission and has triggered the development of the immensely beneficial theories of communication and information transfer.

Important directions for research with significant promise for the future include the following:

In addition, the panel saw overwhelming evidence of the need to couple these techniques with other aspects of system design and technology that are used in the lower and higher layers of the networking architecture. Thus, space-time coding for multiple antenna systems, hybrid error control methods, joint routing/link-layer design, and energy efficiency are examples of integrated approaches that will be necessary for future high-performance, ultra wide band systems.

In other words, again, the panel confirmed the need for vertical integration.

Switching/Networking

In the area of switching/networking, the panel observed the current incompatibility between the available architectures and standards, and the overwhelming attention paid to third generation system design. However, even through the third generation, the air interfaces, the signaling methods, and the handling of roaming will remain dissimilar and incompatible. The third generation migration of GSM includes a packetized transfer mode (referred to as "general packet radio service" or GPRS). However, the incorporation of Internet Protocol (IP) in wireless networks is fraught with problems. Perhaps a sure way to achieve a short term convergence between wireline and wireless networks is to rely on a connection-oriented, circuit-switched architecture.

The key obstacles (and, hence, research areas) that make this aspect of wireless systems more uncertain and less developed are the lack of understanding of how wireless access impacts the definition of a scalable architecture, the lack of suitable protocols for interoperability between the core network and its wireless extensions, and the difficulty of tracking mobility in real time (mobile IP is not a true mobility tracking protocol).

Some notions that show promise for resolving these thorny issues involve the expansion of the "intelligent network" (IN) concepts, the imaginative use of distributed databases and algorithms, and, perhaps, a multi-tier architecture that will allow expansion of scale and will permit new services not yet developed or even defined. Ultimately engineers will need to take a hard look at the possibility of end-to-end IP-based radio systems.

Integrated Design

In the area of "holistic," or vertically integrated design, which emerged as a sufficiently dominant concept to generate a separate section in the report, the major findings were, as already mentioned, that the development of software-defined radios and energy/power management techniques were of foremost interest and immediate concern.

Software radios are seen, in their early stages, as simply reconfigurable, multimode/multiband radios. However, the extended concept of a software radio involves a wide front end and a totally software-dependent baseband architecture that is capable of configuring the operation of the equipment to conform to almost any desired protocol, algorithm, or processing structure. Although the realization of such a concept, the most futuristic version of which anticipates downloading to the terminal the code that will define its operation in real time, is still years if not decades away, its potential is formidable. Assuming questions of cost can be resolved (through economics of scale) every mobile user can customize his own terminal to its current environment. And, if questions of cost are not resolved, software radios can reside only at the base station so that, with their superior "smartness" characteristics, they can adapt their communication to the diverse and simple individual terminals on a one-to-one basis.

Energy savings can be realized in a variety of ways all along the stack of network protocols. From improved battery technology and more efficient chip layouts, to better algorithms for signal processing and adaptive antenna arrays, to improved coding/modulation techniques, to power management techniques in media access control, numerous avenues are available to increase the longevity of untethered terminals and equipment and to save operational and thermal costs at base stations. What is new is the realization that in addition to these methods, network protocol design can have a significant impact on energy efficiency. From pulsed operation of batteries (as for example in time division multiple access-TDMA-schemes) to intelligent routing and multicasting, network protocols can create substantial energy savings. Furthermore, in addition to transmission energy expenditures, there is energy consumed for processing and for simply being "on." Thus, schedules of alternating between "sleep" and "on" modes (as in current paging systems), joint design of compression and transmission schemes, etc., provide means for better overall energy efficiency.

A SAMPLING OF TECHNOLOGY "HIGH-POINTS"

Since this study was performed as an "assessment" of current and future technology, and since its conduct involved numerous site visits around the world, it is appropriate to summarize a few of the dominant points of high technology that the WTEC panelists observed.

The widely espoused notion of "software radio" was a common theme throughout the site visits. The panel was struck by the variety of levels at which this notion is conceived in different quarters. From the simplest version of multi-mode/multi-band radios presented at Ericsson to the most imaginative versions encountered elsewhere, it was clear that this will be the way of the future.

The integration of the antenna into the handset, seen at Nokia, was another harbinger of the direction of future development.

The panelists even saw silver-plated sets that permit the installation of the antenna inside the terminal through a thin-strip plastic window. A great deal of design effort was needed in shaping the antenna plate.

The availability of polymer batteries, displays, and electronics, seen at Philips, is another harbinger of things to come. In addition to reducing weight and costs, these promise superior performance. The first integrated channel characterization models, seen again at Philips, combine electromagnetics, statistics, and computation to provide useful and reasonably accurate models of channel behavior.

The packaging techniques at Matsushita were clearly indicative of superior achievement in the hardware aspects of wireless communication.

The development of three-dimensional integration at Nippon Telegraph and Telephone (NTT) clearly points the way toward size and weight reduction as well as improved performance.

At many sites (especially NTT) the panel was impressed by the effort being made toward reducing the cost of phased arrays.

The SKYNET concept outlined at Yokosuka Research Park (YRP) in Japan is an impressive example of imaginative wireless networking. It is based on a small number of large balloons, suspended in the stratosphere and kept in place by solar-powered propellers, that permit the installation and support of multiple antennas of various sizes. Through direct link connection to terrestrial sites and users, coverage spans the entire region of Japanese territory. Thus, a constellation of satellite-like base stations creates an adequate and relatively inexpensive infrastructure for wireless connectivity.

Study team members were also impressed by the actual deployment of a multi-user detector that utilizes successive interference cancellation in CDMA systems. Although the theory of multi-user detection is quite mature by now, concerns about cost and implementation have held back the incorporation of advanced detectors in actual systems. At NTT DoCoMo, the developed system was at the stage of actual field trials and tests.

Last, but not least, the panel found the advanced miniature multimedia terminals at NTT DoCoMo very impressive and, also, forerunners of future development.

CONCLUSIONS

It is difficult to summarize the findings, observations, and impressions of such a vast study of such an evolving discipline into a few concrete nuggets. Yet, it became clear to this panel that, after the dust had settled, three major items stood out as the major areas of future technology that require significant additional research. These include the following:

All three stood out as being dominant concerns of all visited companies and institutes, both in the United States and abroad. It is also the unanimous and informed belief of the WTEC study team that resolving these issues is necessary in order to realize the ambitious goal of global wireless connectivity.


Published: July 2000; WTEC Hyper-Librarian