Broadband Wireless Access

Copyright 1997 IEEE. Personal use of this material is permitted. However, permission to reprint/republish this material for advertising or promotional purposes or for creating new collective works for resale or redistribution to servers or lists, or to reuse any copyrighted component of this work in other works must be obtained from the IEEE.

This article was published in the January 1997 issue of
IEEE Communications Magazine.

Technical Paper

Abstract

This article reviews the technologies and potential markets, applications, and architectures for broadband wireless access. The emergence of wireless communications for cellular systems is presented, and its projected future evolution to mobile wideband systems. The field of broadband access systems, services, and network architectures is also covered, and then systems for broadband wireless communications for indoor local area networks and outdoor public fixed access networks are discussed. Broadband wireless access systems are emerging as a new and growing area of telecommunications, since the ability to provide access without extensive installation of copper or fiber infrastructures make wireless technology well suited for broadband services. Finally, some of the key enabling technologies, such as adaptive antennas and video compression, and the future architectural directions of broadband wireless networks are presented.

Broadband Wireless Access

Walter Honcharenko, Jan P. Kruys, David Y. Lee, and Nitin J. Shah, Lucent Technologies

Wireless access systems have gained extensive acceptance for mobile and fixed narrowband communications, such as paging, cellular communications, personal communications services (PCS) [1], with projections of over one to two hundred million subscribers by the year 2000. The intervention of wireless technology to provide instant networks [2] in countries which typically lack existing fiber or copper infrastructure is an example of how wireless access has made the deployment of fixed narrowband telephone services possible, where the cost and time to deploy a conventional network would have been prohibitive.
For a number of years, the extension of mobile systems to broader bandwidths and a richer variety of services than messaging, paging, voice, and low-bit-rate data have been explored in activities around the world. Figure 1 is a summary of the generations of wireless communications systems; the activities for the third-generation systems are, for example, illustrated in the Future Public Land Mobile Telecommunications System (FPLMTS)/ International Mobile Telecommunications by the year 2000 (IMT-2000) documents [3]. These activities have concentrated on overall objectives and migration from today s second-generation mobile communications systems to third-generation systems, and include terrestrial and satellite communications. The specific areas which have been studied, for example, in the UMTS (Universal Mobile Telecommunications Systems) bodies are services (including bearer services, teleservices, and interworking with existing networks), charging and accounting, switching and signaling, and network operations and management. It should be noted that bearer services which qualify as broadband access (i.e., hundreds of kb/s up to 2 Mb/s) are defined in the FPLMTS/IMT-2000 for both fixed and mobile access.
Today, the challenge is to innovate and deploy the wireless access technology that will drive this evolution and maintain the high quality of service that the telecommunications and television industries already provide. There are two classes of broadband wireless access systems with which we are familiar today. One is the broadcast of television, where use of radio spectrum, either with terrestrial radio transmissions or via satellite systems, has provided broadband access to the televised broadcast media. These systems have a range of tens to hundreds of kilometers, and have achieved a penetration rate of billions of residential and business subscribers across the globe. The other form of broadband wireless access today is a very geographically localized technology, the wireless local area network (LAN). These systems have emerged only in the past few years, are targeted primarily for business applications, are completely digital, are mainly aimed at data transactions rather than voice or video, and have proven to be most efficient at connecting portable computing and data entry systems to a local network via a high-data-rate (typically a few megabits per second) geographically confined (typically tens to hundreds of meters) environments within an enterprise.
The technology trend for these systems is that the televised media have been driven into cable distribution systems, so the technology is easily scaled, allowing a larger number of entertainment channels and eliminating the need for large external antenna systems in urban and suburban areas, and migration to satellite systems for both rural or remote applications and to support newer digital formats with small-sized satellite receivers. These systems also promise interactive two-way connectivity for data and internet services. Wireless LANs are migrating from modest data transfer rates of a few megabits per second to much higher rates, with projections of tens and hundreds of megabits per second for future systems, for connectivity to Transmission ControlProtocol/Internet Protocol (TCP/IP)-based Ethernet and asynchronous transfer mode (ATM) networks. These newer systems plan to occupy radio spectrum at microwave and millimeter wave frequencies, where there is spectrum available for services requiring a large amount of bandwidth.
Having captured these two extremes of current examples of wireless broadband networks, we have identified several key parameters: the nature of the applications, the data rates, the range and ubiquity of transmission, and issues of spectrum utilization. Taking these as the present base, this article examines first the services and architectures which constitute broadband access systems in general, largely based on delivery with fiber or copper links to the end user. Then we examine in more detail the areas of broadband wireless access which are being explored now. One is the use of spectrum in the microwave and millimeter-wave bands for fixed access in outdoor, public commercial networks; the other is the evolution of wireless LANs for in-building systems. The third and newly emerging area is the use of a LAN technology for outdoor rather than indoor LAN systems, providing large, data-oriented bandwidths on an outdoor campus or in a similarly localized geographic area.

Broadband Access Systems Today

Figure 2 is an illustration of the public broadband access networks (consisting of servers, a distribution network, and a variety of access network alternatives) [4]. The role of the media servers is to provide storage and connectivity to broadband information and other networks. The service nodes are the gateway into the access and distribution network, and the broadband access system provides connectivity into residences and buildings. In Table 1, it is useful to note that the broadband services to be carried on some of these networks may each have their own traffic characteristics [5] and performance requirements such as delay and bit error rates, as well as different bit rates. Reference [6] shows how fiber networks and wireless networks can be achieved from a single server and network architecture [7] such that wireless and conventional schemes can complement each other in a service deployment.
Debate continues on whether wireless multimedia offerings will begin in the business or consumer sector. The answer may be that growth of wireless multimedia will be driven mainly by Intranet and Internet access, and will happen in both the business and consumer sectors simultaneously. The attraction of wireless networking is the ability to provide connectivity rapidly with minimal infrastructure, and to support an initially sparse subscriber base with low penetration rates. Furthermore, if either mobile or portable service is required, wireless technology offers unique capabilities over conventional systems.
The access network is the key to the realistic deployment of broadband services. For conventional access, the critical technologies are integrated services digital network (ISDN) basic rate interface (BRI), ADSL, SDSL, and VSDL (asymmetric, symmetric, and very-high-rate digital subscriber line). These offer data rates ranging from 1.5 Mb/s to 51 Mb/s for the highest rates. Hybrid fiber coax systems, which allow flexible transport of analog cable television signals as well as digital formats and support of data links over the same medium using frequency division multiplexing (or pass-band technology) with cable modems, are yet another technology solution. These systems utilize coaxial cable for the final link to the end subscriber, fed by optical fiber, similar to existing cable TV networks, but supporting analog and digital two-way services, unlike conventional cable broadcast networks. Switched digital video systems (also referred to as fiber to the curb systems) also provide digital broadband access for interactive and broadcast television systems. These systems use purely digital access, with baseband digital information transmitted by optical fiber to the curb, and a small-bandwidth copper connection to homes or businesses (typically tens of residences would be supported by one optical fiber termination). In the future, the technology trajectory for many of these systems is to eliminate much of the electronics and optical elements in the outside plant equipment, and to have passive optical networks, which offer the highest reliability and deployment ease for the network operator.

Service Quality in the Wireless Environment

Unlike transmission in the wireline environment, where it is possible to increase the available transmission spectrum by using another wavelength (wavelength division multiplexing) or installing another cable, wireless transmission is limited by available radio spectrum and impaired by both interference and multipath propagation, which cause fading and delay spread. Because of these limitations wireless networks, in general, will have lower data rates and higher error rates than comparable wireline networks for economical deployment. However, technology can enable network operators to provide customers with wireless access to services typically associated with broadband wireline facilities.
Figure 3 maps telecommunications services (ranging from voice to video on the left) to the nature of the traffic (ranging from circuit- to packet-switched on the right) and the associated bandwidths (on the bottom) to a number of applications (in the center). With wireless systems as the transmission medium, the typical cellular and personal communications services (PCS) systems today reside at typical use data rates of tens of kilobits per second in wide-area coverage, served by macrocells and microcells. Wireless LANs, with from 2 to over 50 Mb/s transmission rates and cell coverage of tens of meters, are also mapped into this chart. Third-generation mobile systems cover many of the dimensions of this chart, but aspire to provide broader-band services (a few tens to hundreds of kilobits per second) in a mobile environment. The fixed wireless environment is likely to support higher transmission rates and broader bandwidth services even in outdoor environments compared to mobile networks, and promises a wide range of future services, including both entertainment and packetized data services such as Internet and Intranet connectivity to homes and businesses.
Table 2 provides a comparison of the spectrum allocations, typical data rates, cell capacity, and ranges for indoor, fixed outdoor, and mobile outdoor wireless access systems, and complements the data in Fig. 3.

Wireless Local Area Networks

In the majority of applications wireless LAN networks are used for extension or replacement of wired LANs: networks of computers and workstations that share processing capacity and filing services. These systems are characterized by bursty traffic patterns between the LAN stations. The messages are typically short on the order of a few hundred bytes but the message frequency can be quite high, as much as hundreds per second. Wireless LAN products were pioneered in the late 1980s and early 1990s to allow wireless connectivity to LANs in buildings for portable computing devices [8, 9].
Spectrum allocation for LANs was initially in the Industrial, Scientific, and Medical (ISM) bands at 900 MHz. Subsequently, spectrum at 2.4 GHz has also been utilized for these systems. The role of the IEEE 802.11 standard was to provide specifications for the wireless LAN industry and to drive widespread deployment of these systems.
The main impetus for this market comes from users who need mobility and portability in their daily work. This requires systems with sustained transmission and coverage across the coverage area, and also support of handovers, making connectivity to the LAN transparent to the user. Since the Ethernet-based packet data transfer supported by these systems is connectionless, handover management is easier to achieve in data networks than voice systems. The applications served are either vertical (e.g., warehousing, medical, or retail locations) or horizontal (e.g., office automation, access to e-mail, or file transfers), and both have grown tremendously in the past few years. In the future, we may see locations where all the local data and voice traffic (in the form of packetized voice) is carried over such networks, and communications are carried out with portable or wearable devices capable of computing as well as voice and data transmission. Wireless LAN networks are frequently used to provide roving users access to wired LAN systems. The speed of movement is typically restricted to walking speed or indoor vehicle speed: a maximum speed of 10 m/s typically covers most in-building applications.
Carnegie Mellon University is a showcase of a living wireless campus, and a research testbed for applications, services, protocols, and so on [10]. This is an example of ad hoc wireless networking. Such networks are most likely to occur in education and training environments where computers are shared between many different people but also in collaborative environments such as team problem solving, collaborative project groups, intra- as well as intergovernmental meetings, and so on.

Wireless ATM

As the deployment of wireless LANs grows, there is a push to higher data rates. As a result, spectrum has been allocated for the high-performance LAN (HIPERLAN) and SUPERNet activities at 5 GHz, and these will support connectivity of 20 to 25 Mb/s. Moving to even higher frequencies, such as 40 and 60 GHz, has also been proposed, where LANs approaching 100 Mb/s are the subject of current research. Although these higher frequencies offer large amounts of spectrum, the radio systems typically require directional antenna systems for robust links. Therefore, they are more suited to fixed links, and the use of the 900 MHz, 2.4 GHz, and 5 GHz bands will continue to have broader applicability.
Because of the wide range of services supported by ATM networks, ATM technology is expected to become the dominant networking technology in the medium term for both public infrastructure networks and LANs. ATM infrastructure can support all types of services, from time-sensitive voice communications and desktop multimedia conferencing to bursty transaction processing and LAN traffic. Extending the ATM infrastructure with a wireless access mechanism [11 15] meets the needs of those users and customers who want a unified, end-to-end networking infrastructure with high-performance, consistent service characteristics. Wireless ATM adds the advantages of mobility (i.e., cordless operation) to the service advantages of ATM networks.
A wireless access point connects the set of wireless nodes it services on a single port of the ATM switch, as shown in Fig. 4. This system could represent either the topology of an Ethernet system for an Intranet, with TCP/IP transport or ATM. The system consists of wireless access points geographically distributed in the building(s) to be covered, and a wired ATM network for connectivity to the wireless access points.
Several design issues arise with the need to transport ATM packets over a radio link. The IEEE 802.11 standard specifies link turnarounds on the order of hundreds of microseconds, far more than the duration of an ATM cell, even at the modest bit rate of 25 Mb/s. The efficiency of encapsulated ATM is therefore low, too low to be practically acceptable. Another problem with this approach is that the existing protocols have no means of communicating load and rate information between nodes. In order to service its nodes in the most efficient manner, the wireless access point would need to know the instantaneous load conditions of its nodes. Wireless protocols like IEEE 802.11 and European Telecommunications Standards Institute (ETSI) HIPERLAN do not support the transfer of such knowledge. A technique called native mode ATM overcomes these limitations. The unit of information transfer is the ATM cell, and the wireless access point functions as an extension of the switch that services the nodes in a sequence best fitted to their advertised traffic needs. The wireless subsystem operates on ATM cells and thus allows existing as well as new ATM-based communications software to be used. This integrated approach also simplifies resource and capacity management.
As an illustration of the capacity projections for a wireless LAN, consider the following scenario: A group of users share a service area (e.g., a floor of a building). The local computing system is a set of servers wired together via fiber distributed data interface (FDDI) and an ATM user premises switch. All users have notebook computers, most have voice capability, and 20 percent have video capability as well. Furthermore, assume that typically one in six of the users are involved in a voice connection or voice plus data connection, whereas only 1 in 20 is involved in a video connection. The following figures are based on a wireless range of 20 m:

Data throughput per service area: 6Mb/s
Voice throughput per service area:640 kb/s
Video throughput:6 Mb/s
Total bit rate / service area:12.6 Mb/s

This simplified scenario shows that wireless LANs with capacity of 10 to 25 Mb/s per access point service area would be sufficient for a system of this kind.

LMDS and MMDS

The local multipoint distribution service (LMDS) and multipoint multichannel distribution service (MMDS) depicted in Fig. 5 represent two approaches being discussed by cable television (CATV) companies, regional telecommunications companies, and others interested in gaining local access to residential areas.

MMDS

The MMDS ( wireless cable ) has existed for some time and offers a maximum of 33 analog video channels in the spectrum between 2.150 and 2.682 GHz at abandwidth of 500 MHz. MMDS systems offer some competition to conventional cable systems, and also provide complementary service in regions where deployment of cable systems is impractical [16]. The typical cell radius of the MMDS extends approximately 25 to 35 mi depending on terrain and antenna placement. Current CATV infrastructure supports only 30 to 60 channels. Satellite systems support about 150 to 200 digital broadcast channels. Thus, current MMDS operators are looking to use digital compression techniques to increase the number of channels to around 200, making the systems more competitive with wired cable services and satellite-delivered programming.
Reference [17] provides many of the basic facts about MMDS systems, including trials of the newer digital technologies which promise higher range and spectral efficiency than current commercial systems. This article also contains information about security, range, and susceptibility to climatic effects which, even at the relatively low frequencies of 2.5 GHz, result in some interruption of service in certain regions with rain and fog conditions. With the evolution of digital systems typically using Motion Pictures Expert Group version 2 (MPEG-2) and source coding for signal integrity and complex modulation schemes such as 64-quadrature amplitude modulation (QAM) for spectral efficiency, these systems may also provide two-way connectivity and transport to the transmitter using ATM or synchronous optical network (SONET) networks from the video server/headend network to the transmission nodes, perhaps using TCP/IP protocols. Furthermore, these systems can support both single and multiple dwelling units with local cable distribution of the received signals to the individual subscribers, such as in an apartment building.

LMDS and LMCS

Local multipoint communications systems (LMCS and LMDS are synonymous) have recently been introduced to deliver broadband wireless access [18, 19]. The LMDS, having cell radii shorter than five mi, can deliver two-way high-speed data, broadcast video, video-on-demand services, and telephony to residential areas. The proposed spectrum allocation for LMDS is between 27.5 and 28.35 GHz, with more spectrum between 29.1 and 29.25 GHz available in areas highly coordinated with fixed satellite links, which also operate in the 27 30 GHz band. LMDS can offer two-way wireless services, whereas MMDS and satellite systems require terrestrial wired networks to communicate back to the headend, for example, to select programming or use VCR-type controls on video-on-demand programming.
The main advantages of both the MMDS and LMDS are similar to those of all wireless systems: ease and speed of deployment. Regional telecommunications companies are finding deployment of cable and fiber systems difficult in certain areas where installing in-ground infrastructure is undesirable (due to existing buildings and infrastructure in place), impractical (due to terrain), and costly (since extensive build-out of the infrastructure is required before commercial service can be launched). LMDS can provide similar access bandwidths and a two-way capability without trenching streets and yards.
The primary disadvantages of both the LMDS and MMDS are co-channel interference from other cells and limitations on coverage. Currently, for example, the Federal Communications Commission (FCC) requires a 35-mi protection zone between MMDS operators to reduce any chance of interference. Coverage issues are not as great a challenge with MMDS as they are with LMDS. Millimeter-wave radio signals do not penetrate trees. Thus, line-of-sight propagation paths are required. This requirement can make antenna placement on subscriber homes challenging. Contrary to intuition, rain fading effects (channel fading due to heavy downpours) for small (< 1 km) cells are minor; however, these effects become more hostile on longer path links, which increase the service outage probability due to a passing rainstorm. The regions of rainfall characteristics have been characterized throughout the world, and there are locations that are particularly devastated by this type of climatic condition.
The most challenging area of these systems, required to operate at these high frequencies, is that even if the transmitter and receiver are located at fixed points, the influence of motion of traffic and foliage, even in a line of sight location, creates a fading environment which is much more hostile than measured for conventional cellular mobile systems at, say, 2 GHz. Temporal fades of over 40 dB at the rate of at most a few (< 2) hertz are frequently seen in these environments (Fig. 6), imposing very stringent requirements on the error correction coding of the transmitted bitstream. A number of studies indicate that the excess path loss and fading characteristics, even in a relative static environment, need extensive characterization before these systems can be made commercially viable [20 25]. Recently, a study for LMDS was performed [26], where the conclusions indicated that excess path loss was the major propagation impediment for proposed LMDS services. However, delay spread could be confined to about 10 ns by the use of narrowbeam sectorized antenna systems. These results will help in the design of a robust LMDS system, and techniques such as overlapping cells and polarization techniques for intersector and cell-to-cell isolation have been proposed to overcome these limitations. The attractiveness of almost 1 GHz of spectrum for broadband wireless access is driving these technology developments for fixed systems, despite the technical hurdles of propagation. Providing acceptable real-time, low-delay, constant-bit-rate services such as video transmission will be a challenge. Innovations in the area of cell design, antenna sectorization, and compression techniques, combined with coding and modulation techniques will be required to overcome the hostile fading and environmental propagation characteristics of this frequency range.

Enabling Technologies

Antenna Systems for Broadband Wireless Access

Antenna systems and techniques for transmission of information at increasingly higher bits per second per hertz in a fixed environment, maintaining low bit error rates, have led to some fundamental studies on the absolute limits that can be attained. This work has formed the basis of the questions and potential theoretical and practical techniques for the solutions [27, 28], including adaptive equalization, diversity at the transmitter and the receiver, and several combining techniques from the reception of multiple antenna systems to define the fundamental limits of maximum throughput with target bit error rates in Rayleigh fading environments. Reference [29] is a general tutorial of techniques applied to adaptive antenna arrays to improve the system performance of wireless networks. Many of the techniques described will have to be commercialized to realize reliable broad wireless systems.
For wireless LANs, although propagation in an indoor environment does not suffer the mobility and characteristics of the outdoor environment, buildings offer clutter and path loss which cause delay spread and multipath. The delay spreads, combined with transmission rates of many megabits per second, requires equalization techniques to ensure good signal quality [30]. Techniques for coding [31] and modulation schemes [32] reduce peak power requirements for RF components, which become more expensive as the frequency of operation and transmitted power levels increase.

Broadband Multimedia Services and Compression/Coding Technologies

Along with the development of radio communications systems and protocols for the transmission of broadband wireless systems, the ongoing technology development for compression and coding of multimedia content is going to facilitate robust services in an inherently lossy transmission environment. Whereas the techniques for non-real-time services (such as automatic repeat request and retransmission schemes) are well established for robust data transport, continued technology innovation is critical for successful delivery of multimedia services.
Reference [33] guides the reader through the definitions of multimedia and many of the gating parameters which govern both the acceptability and quality of services. A summary of relevant coding algorithms and standards for image, speech, and video services are provided, and the associated system parameters, such as processing delay, quality, and algorithmic complexity (which often affects the cost and viability of the fixed or mobile end-user equipment containing the processors to perform the coding and decoding functions).

Conclusions

Broadband wireless access systems have the potential to deliver services with the quality and availability of conventional networks, yet offer the potential for rapid deployment characteristic of wireless networks. As regulatory changes make more spectrum available, the use of wireless LANs and microwave and millimeter-wave frequencies for fixed wireless systems will become more attractive. Numerous technical challenges must be overcome, and services that offer unique characteristics over conventional networks will emerge.
Distributed network architectures which support a variety of wired and wireless access schemes for both mobile and fixed services and for both narrowband and broadband traffic (Fig. 7) will emerge to provide the backbone network architectures for these new systems. These systems will also support the very high data rates described above, not only for in-building or residences, but also for campus and open spaces, where portable communications devices will be able to gain access to broadband data, internet, and multimedia services.

References
[1] T. S. Rappaport, B. D. Woerner, and J. H. Reed, Wireless Personal Communications: The Evolution of Personal Communications Systems, Boston: Kluwer, 1996.
[2] G. I. Zysman, Wireless Networks, Scientific American, Sept. 1995, pp. 68 71.
[3] ITU-R Rec. M.687, Future Public Land Mobile Telecommunications Systems, and ITU-R Recs. M.1034 and M.1035, Requirements for the Radio Interface(s) for FPLMTS.
[4] W. Pugh and G. Boyer, Broadband Access: Comparing Alternatives, IEEE Commun. Mag., vol. 33, no. 8, Aug. 1995.
[5] T. J. Aprille et al., End to End Consumer Interactive Networks, submitted to the Int l. Symp. on Subscriber Loops and Services 96, Melbourne, Australia, 1996.
[6] N. S. Stevens, T. J. Aprille, and B. D. Bolliger, Integrated Fiber and Wireless Access Systems, submitted to the National Fiber Optic Engineers Conference, Denver, CO, 1996.
[7] T. J. Aprille, P. DeDuc, and W. Kremer, SONET Rings, Ready to Face the Future, America s Network, vol. 98, no. 24, Dec. 15, 1994.
[8] B. J. Tuch, An ISM Band Local Area Network, WaveLAN, Proc. IEEE Wksp. on Local Area Networks, Worster Polytechnic Institute, Worster, MA, May 1991, pp. 103 11.
[9] J. Kruys, HiperLAN, Applications and Requirements, Proc. PIMRC 92, Boston, MA, Oct. 1992, pp. 133 35.
[10] M. Satyanarayanan, Mobile Computing at Carnegie Mellon, IEEE Pers. Commun., vol. 3, no. 1, Feb. 1996.
[11] M. Schwartz, Network Management and Control Issues in Multi-media Wireless Networks, IEEE Pers. Commun., June 1995.
[12] D. Raychaudhuri and N. D. Wilson, ATM Based Transport Architecture for Multiservice Wireless Personal Communications Networks, IEEE JSAC, vol.12, Oct. 1992, pp.1401 14.
[13] K. Y. Eng et al., A Wireless Broadband Ad-Hoc ATM Local Area Network, Wireless Networks, 1995, pp. 161 74.
[14] P. Agrawal et al., SWAN, A Mobile, Multimedia Wireless Network, IEEE Pers. Commun., Apr. 1996.
[15] Magic WAND, available at http://www.tik.ee.ethz.ch/~wand.
[16] G. Lawton, Competition Comes to the Cable Industry, Commun. Tech., Sept. 1995, pp. 24 29.
[17] B. J. Catlin, Wireless Cable Television: Frequently Asked Questions.
[18] D. A. Gray, Broadband Wireless Access Systems at 28 GHz, CED, July 1996, pp. 46 56.
[19] F. Dawson, Cellular TV Adds New Wrinkle to War, CED, Sept. 1996, pp. 98 102.
[20] A. A. Ali and M. A. Alhaider, Effect of Multipath Fading on Millimeter Wave Propagation: A Field Study, IEE Proc.-H, vol. 140, no. 5, Oct. 1993.
[21] C. E. Hendrix et al., Multigigabit Transmission through Rain in a Dual Polarization Frequency Reuse System: An Experimental Study, IEEE Trans. Commun., vol. 41, no. 12, Dec. 1993, pp. 1830 37.
[22] P. B. Papazian, D. L. Jones, and R. H. Espeland, Wideband Propagation Measurements at 30.3 GHz through a Pecan Orchard in Texas, NTIA Rep., 1991.
[23] R. P. I. Scott, 29 GHz Radio System for Video Distribution, BT Tech. J., vol. 8, no. 4, Oct. 1994.
[24] F. K. Schwering and E. J. Violette, Millimeter-Wave Propagation in Vegetation: Experiments and Theory, IEEE Trans. Geoscience and Remote Sensing, vol. 26, no. 3, May 1988, pp. 355 67.
[25] E. J. Violette et al., Millimeter-Wave Propagation at Street Level in an Urban Environment, IEEE Trans. Geoscience and Remote Sensing, vol. 26, no. 3, May 1988, pp. 368 80.
[26] P. Papazian and G. Hufford, Initial Study of the Local Multipoint Distribution Service Radio Channel, Proc. Wireless 96 Conf., Calgary, Alberta, Canada, July, 1996, pp. 494 513.
[27] J. H. Winters, On the Capacity of Radio Communications Systems in a Fading Environment, IEEE JSAC, June 1987, pp. 871 78.
[28] G. J. Foschini, Layered Space-Time Architecture for Wireless Communications in a Fading Environment Using Multi-Element Antennas, Bell Labs Tech. J., Dec. 1996.
[29] A. Paulraj, Third Workshop on Smart Antennas in Wireless Mobile Communications, Ctr. for Telecommun. and Info. Sys. Lab., Stanford Univ., Stanford, CA, July, 1996.
[30] B. Daneshrad, The HIPERLAN Equalizer ASIC Complexity and Its Relationship to the Training Header, accepted for publication in Wireless Pers. Commun., 1996.
[31] T. A. Wilkinson and A. E. Jones, Minimization of the Peak-to-Mean Envelope Power Ratio of Multicarrier Transmission Schemes by Block Coding, Proc. IEEE Vehic. Tech. Conf., Chicago, IL, July 1995.
[32] A. Kameran and A. S. Krishnakumar, OFDM Encoding with Reduced Crest Factor, Proc. 2nd Symp. Commun. and Vehic. Tech., Benelux, Belgium, Nov. 1995.
[33] N. Jayant et al., Multimedia: Technology Dimensions and Challenges, AT&T Tech. J., vol. 74, no. 5, Sept. Oct. 1995, pp. 14 33.

Biography
Walter Honcharenko joined the Wireless Core Technology Department of Lucent Technologies (then AT&T) in 1993 as a member of technical staff. Since joining, he has been involved with cellular and PCS propagation studies, adaptive antennas systems, and millimeter-wave broadband access systems. His current responsibilities include micro/millimeter-wave broadband wireless access. Dr. Honcharenko received his Ph.D. in electrophysics (1993), and M.S. (1991) and B.S. (1989) in electrical engineering, all from Polytechnic University, Brooklyn, New York. E-mail:honcharenko@lucent.com

Jan Kruys joined Lucent Technologies WCND (then NCR) in 1971. Since then he has worked on communications hardware and software design, on-line payment systems, distributed systems security, wireless LANs, and access networks. He served as representative with ECMA, ANSI, ETSI and WINForum. His current responsibilities include chairing ETSI RES 10 and leading Lucent Technologies contributions to the Magic WAND consortium (ACTS, Europe). E-mail: kruys@lucent.com

David Y. Lee is a technical manager in the Wireless Systems Core Technology Department in the Wireless Technology Laboratory at Lucent Technologies. He is responsible for wideband (5 MHz) CDMA prototyping, base-station technologies, as well as management of intellectual property, technology planning, and technical marketing. Mr. Lee holds a B.A. degree in chemistry and a B.S. in electrical engineering from the State University of New York (SUNY) in Stony Brook. He also has an M.E.E. degree from Cornell University in Ithaca, New York. E-mail:dylee1@lucent.com

Nitin J. Shah is technology director of the Wireless Core Technology Department in Lucent Technologies. His group is part of the network infrastructure manufacturing division of Lucent Technologies, responsible for cellular, personal communications networks, wireless local access, and other wireless communications products. He has projects on technology planning, wireless network architecture, radio multiple access technologies, and digital compression technologies for speech and visual communication. Dr. Shah received his B.A. (1979), M.A. (1982), and Ph.D. in microelectronic engineering (1983) from the University of Cambridge, England. E-mail:njshah@lucent.com