FIXED TERRESTRIAL BROADBAND WIRELESS ACCESS



-FIXED BROADBAND WIRELESS ACCESS

An update of TTG-3 Report [1]

Ferdo Ivanek

MTT-S Representative to CCIP

Ivanek@stanford.edu

1. Introduction

The objective of this update of the TTG-3 Report [1] is to evaluate the wireless alternatives to wireline broadband access provided over cable TV networks, local telephone networks (e.g. DSL, T1), and direct fiber connections to the end user. Developments in the U.S. are only covered in accordance with the IEEE-USA framework.

The introduction starts with terminology and background. Section 1.1 covers the basic broadband wireless access (BWA) implementation options, and Section 1.2 describes the BWA framework of the IEEE 802 LAN/MAN Standards Committee. Section 1.3 draws attention to the “fixed wireless continuum” that includes MAN (metropolitan access network), LAN (local area network), PAN (personal area network) and optical wireless access. Sections 2 and 3 cover fixed BWA in frequency bands above 20 GHz and below 6 GHz, respectively, which account for the predominant applications in the U.S. Section 4 addresses free-space optical access, and Section 5 presents the conclusions and comments.

Terminology - Broadband wireless access, or BWA, is the most widely used generic term for subscribers’ “first-mile”/”last-mile” terrestrial fixed broadband wireless access to the telecommunications network. More specific terms are used in regulatory and standardization contexts. The FCC established BWA rules under different names for different frequency bands, such as, Multipoint Distribution Service (MDS), Local Multipoint Distribution Service (LMDS), and Unlicensed National Information Infrastructure (U-NII). Within the IEEE 802 LAN/MAN Standards Committee, WirelessMANTM designates the family of standards for BWA systems, where “MAN” stands for “metropolitan area network”.

Background - Large scale fixed BWA deployment in the U.S. was stimulated by the Telecommunications Act of 1996 and expanded rapidly until 2001, when it suffered a serious setback under the impact of a combination of adverse conditions. Nevertheless, the post-crash landscape shows achievements of lasting value and a promising level of continuing activities, for example:

• the best established BWA service, deployed in the 39 GHz band, emerged from Chapter 11 protection under new ownership and continues regular operation;

• the downsized systems supplier community is intensifying work on the next generation(s) of BWA systems with greatly improved capabilities at lower cost;

• FCC proceedings are in progress for new BWA applications in the 70 GHz, 80 GHz and 90 GHz frequency bands;

• BWA standardization within the IEEE Standards Association is in progress with widespread industry participation.

1.1 BWA implementation options

The BWA implementation options include:

• licensed and unlicensed operation in a number of frequency bands allocated to the fixed and mobile services, as well as free-space optical operation;

• access link lengths ranging from hundreds of meters to over 20 kilometers, depending on the propagation characteristics that are determined by the choice of frequency band and path, and by the local meteorological conditions;

• point-to-point (P-P) systems applications in star/cellular and ring topologies, point-to-multipoint (P-MP) systems in cellulatr topology, and multipoint-to-multipoint (MP-MP) systems in mesh topology;

• the predominant transmission formats and protocols: SONET/SDH, ATM, IP and Ethernet;

• frequency division multiple access (FDMA) and time division multiple access (TDMA)

with frequency division duplex (FDD) or time division duplex (TDD) operation;

• the use of advanced signal processing techniques for efficient spectrum utilization, countermeasures for electromagnetic wave propagation impairments, and privacy;

• quality of service (QoS) ranging from ATM equivalent to best effort.

The multitude of options make it possible to implement a variety of BWA systems optimized for specific applications.

1.2 The IEEE standardization framework

Recognizing the BWA potential, the IEEE 802 LAN/MAN Standards Committee established the IEEE 802.16 Working Group on Broadband Wireless Access. Work is in progress on a family of standards intended to enable multiple vendors to produce interoperable P-MP and MP-MP mesh BWA systems for a number of frequency bands in the 2 - 66 GHz frequency range. Significantly, this is the first U.S. standardization effort in fixed terrestrial wireless access systems, complementing the work of IEEE 802.11 Working Group on Wireless Local Area Networks. IEEE 802.16 activities are highly productive, they are transnational, greatly influence international standardization efforts, and substantially improve the overall BWA outlook.

IEEE 802.16 developed a common medium access control layer (MAC) and separate physical layers (PHY) for the 10 - 66 GHz and 2 - 11 GHz frequency ranges, respectively. This differentiation takes into account the facts that the higher frequency range encompasses an order of magnitude more spectrum available for BWA applications but is usable only for line-of-sight (LOS) operation, whereas near-line-of-sight or non-line-of-sight (NLOS) operation is feasible only in the lower frequency range. Sections 2 and 3 reflect this fundamental distinction and focus on the frequency ranges above 20 GHz and below 6 GHz, respectively, which encompass the major BWA bands.

3. The fixed BWA panorama: MAN, LAN, PAN and optical wireless access

This draft adopts the IEEE 802.16 definition of BWA, which covers only MAN. However, optimal MAN development requires taking into account related developments within the “fixed wireless continuum” which encompasses MAN, LAN and PAN, as well as optical wireless access. IEEE 802 addresses LAN in 802.11 and PAN in 802.15. In this broader context, BWA complements and competes not only with wireline alternatives but also with mobile broadband wireless alternatives. The latter case is best exemplified with the fast growing wireless LAN deployments that already satisfy widespread “nomadic” fixed access needs targeted by third generation (3G) or fourth generation (4G) mobile systems (e.g. 2 Mb/s) at some future time. But 3G/NG will only be able to accomplish such needs by making inefficient use of their capabilities optimized for mobile operation.

1. BWA in frequency bands above 20 GHz

Since propagation impairments generally worsen with increasing frequency and thereby shorten the usable path lengths, frequency bands above 20 GHz are particularly well suited for “first-mile/last-mile” subscriber access connections. Only line-of-sight (LOS) paths are usable in this frequency range.

The acronym LMDS, which stands for Local Multipoint Distribution Service, commonly applies to BWA applications in these frequency bands.Originally, LMDS was planned to be a one-way wireless substitute for cable TV in the 28 GHz band. It has not materialized as planned, and this band is now targeted for BWA applications. However, substantial deployment of this kind of service started first in the 39 GHz band, and then in the 24 GHz band, both by CLEC’s providing LOS access to office buildings in urban areas. Initial deployment used P-P links in cellular network configurations. P-MP systems were introduced more recently. Subscriber densities currently range up to about 100 stations per square kilometer, and provide substantial room for further growth. Data rates ranged initially from T1 to T3, and more recently to OC-3 for P-P systems and 200 Mb/s for P-MP systems. These BWA networks are typically designed to match the bit error rate performance of fiber access and to achieve 99.999% availability. P-P systems carrying OC-12 have been introduced recently, as well as a Gigabit Ethernet system for the unlicensed 60 GHz band.

More widespread deployment of such BWA systems is facilitated by a WRC-2000 decision and by IEEE 802 LAN MAN standardization. WRC-2000 made available several bands in the 30 - 66 GHz frequency range for “high density applications in the fixed service”, whereas the IEEE 802.16 Working Group developed the IEEE 802.16 WirelessMANTM air interface standard for P-MP systems in the 10 - 66 GHz frequency range. Maturing technology for ever higher frequencies continues to shift the limit of cost-competitive BWA applications upwards. As a consequence, FCC proceedings are in progress for new applications in the 70 GHz, 80 GHz and 90 GHz frequency bands, and industry is already aiming at even higher frequency bands. These developments open the prospect of large scale BWA deployment at ever higher data rates that are increasingly more competitive with both fiber cable and free-space optical access up to multi-Gb/s data rates. As a near-term example, a 10 Gigabit Ethernet system operating in the license-exempt 57 - 64 GHz frequency range is announced to become available in 2003.

3. BWA in frequency bands below 6 GHz

The general trend of propagation conditions, pointed out in the first paragraph of Section 2, makes the bands below 6 GHz particularly desirable for BWA applications requiring longer LOS subscriber access links and/or denser service penetration through the use of NLOS links. However, the frequency spectrum available for BWA applications in this frequency range is more than an order of magnitude smaller than above 20 GHz, and is partially targeted for mobile usage.

The major frequency bands below 6 GHz that are available for BWA in the U.S. are the 2.5 GHz band licensed for the Multichannel Multipoint Distribution Service (MMDS), and the 5 GHz bands for the Unlicensed National Information Infrastructure (U-NII). Both types of applications are covered by the IEEE 802.16a WirelessMAN standard under development.

3.1 MMDS

BWA is a more recent application in the 2.5 GHz band originally intended for “wireless cable TV” sharing with the closed circuit Instructional Television Fixed Service (ITFS). ILECs lacking wireline local access facilities were the principal BWA users, so far, providing wireless alternatives to DSL and Cable TV modems. The initial deployment pattern using first-generation MMDS systems was similar to that of LMDS, but the more favorable propagation conditions in the 2.5 GHz band made it possible to deploy substantially larger cells in LOS operation, as well as to operate near-line-of-sight links. Non-line-of-sight (NLOS) operation is being planned using substantially improved and more versatile second-generation MMDS systems, which promise to substantially increase service penetration in comparison with that of LOS operation. NLOS operation is also expected to facilitate self installation by the subscriber, particularly with indoor mounted antennas. However, NLOS imposes a trade-off between increased service penetration and lower achievable availability in comparison with LOS operation. Nevertheless, the typical 99.9% availability for NLOS operation is considered to be acceptable to a sizeable subscriber population if pricing is competitive.

First generation MMDS systems turned out to be too costly in competition with the cable TV and DSL alternatives. This brought MMDS deployment to a temporary halt, awaiting the second generation P-MP and MP-MP mesh systems which promise a number of substantial improvements in deployment, operation and service profitability. These new systems are being evaluated by interested service providers. However, major new MMDS deployment is unlikely before the completion of the IEEE 802.16a standard for the air interface in the licensed bands in the frequency range 2 - 11 GHz. The current draft includes three options: single-carrier, orthogonal frequency multiplexing (OFDM) and orthogonal frequency division multiple access (OFDMA).

In addition to this short-term uncertainty, there is the long-term uncertainty about the continued suitability of the 2.5 GHz band for BWA applications. WRC-2000 identified this band for IMT-2000 use, and the FCC opened it up for shared mobile use but without apparent intent to restrict MMDS and ITFS usage. Market forces may well determine the future usage of this band in the U.S. Much will depend on the preference of the major service providers who are in both the fixed BWA and cellular mobile markets. Significantly, IEEE 802.16 is already considering limited mobility as a near-term objective, and has recently been approached with a proposal to also consider mobile BWA in licensed bands “supporting mobility at vehicular speeds.”

There are also concerns about whether the 200 MHz wide 2.5 GHz band is wide enough for extensive MMDS deployment. One recent proposal is to extend MMDS usage into the adjacent license-exempt 2.4 GHz industrial, scientific and medical (ISM) band. However, the ISM band is already crowded with a number of different users. Wireless LANs based on the IEEE standard 802.11b account for the fastest growing usage of this band. Relief is expected from the new LAN version based on the IEEE standard 802.11a for applications in the U-NII bands.

2. U-NNI

The FCC Report and Order on the Unlicensed National Information Infrastructure produced a uniquely versatile solution for an attractive set of multimedia service needs of a diverse group of wireless users requiring substantially higher data rates than available from MMDS. A total of 300 MHz of spectrum was made available for this purpose in the 5 GHz frequency range, subdivided into three 100 MHz segments differentiated by technical requirements restricting potential interference to licensed and license-free systems sharing the same frequency band. The FCC differentiation includes identification of preferential usage of each of the three bands: 5.150 – 5.250 GHz for indoor LAN, 5.250 – 5.350 GHz for campus-type LAN, and 5.725 – 5.825 GHz for “community networks with a typical range of several kilometers.” This allows to satisfy a great variety of needs with a small variety of systems operating within adjacent or closely spaced frequency bands. These favorable conditions would be enhanced by making the 5.470 – 5.725 GHz band also available for U-NII applications, as has been proposed recently.

The FCC differentiation makes the band 5.725 – 5.825 GHz the most suitable one for BWA applications of conventional P-P systems, or P-MP systems being standardized in the IEEE 802.16a project. However, the license-exempt status of all three bands tends to relax the intended application differentiation. As an example, some 802.11a products cover all three bands. And there are more indications of the tendency to use the license-exempt U-NII spectrum more liberally than differentiated by the FCC. This is facilitated by the lack of rules for efficient spectrum use and coordination among services deployed in the same area. The risk of interference will increase with increasing band usage, and it is in the service providers’ best interest to address this potential problem in some cooperative way. A precedent exists for BWA applications in the licensed bands above 10 GHz, for which IEEE 802.16 has developed IEEE Standard 802.16.2 entitled “IEEE Recommended Practice for Local and Metropolitan Area Networks – Coexistence of Fixed Broadband Wireless Systems”.

3. 2.4 GHz ISM band: the significance of the IEEE 802.11b LAN developments

The rapid growth of LAN deployment in the 2.4 GHz license-exempt band, based on the IEEE standard 802.11b, offers useful lessons not only for the LAN prospects in the U-NII bands, based on the IEEE standard 802.11a, but also for MAN prospects in a number of bands, based on the IEEE 802.16 standard. The single most important lesson seems to be that license-exempt operation and product consumerization are the most effective catalysts for large scale fixed BWA applications, especially when both apply. But this is just another demonstration of the benefits of improving user friendliness in the development of telecommunications services, which are accelerating the “do it yourself” trend among end users, enterprises and communities. Multi-tenant buildings, small- office/home-office (SOHO), and residential wireless MAN applications in either licensed or license-exempt band would similarly benefit from product consumerization and ease of self-installation. Subscribing to a BWA service has the potential of becoming as simple as purchasing a TV set and installing an indoor or outdoor antenna.

4. Free-space optical access

Free-space optical access systems naturally surpass the data rate capability of millimeter-wave BWA systems, and are therefore the most promising competitor to fiber cable access for link lengths that make it possible to achieve 99.999% availability under the local visibility conditions in the presence of fog, smog and mist. However, this availability requirement limits optical link lengths to a greater extent than for millimeter-wave links, except in localities with highest rainfall rates. Some suppliers of free-space optical systems consider lower availabilities, e.g. 99.9%, acceptable to users without a more favorable alternative, and they propose millimeter-wave back-up for “stretched” links requiring 99.999% availability. Both possibilities would improve the market outlook.

5. Conclusions and comments

Prerequisite: cost-competitive technology - The progress of wireless technology, spectrum availability and systems implementations effectively support the deployment of a variety of BWA services for which market demands provide a sound business case. Cost-competitiveness in providing a needed service is the ultimate prerequisite that systems suppliers and service providers are jointly striving to satisfy and improve under persistent market pressures.

Impact of business conditions - The future of fixed BWA depends to a large extent on factors other than technological capabilities to satisfy subscribers’ needs for broadband access. The recent BWA setback provides valuable indications on the negative impact of lacking cost-effectiveness, unintentional negative consequences of regulatory measures, persistent strong competitive pressures from dominant incumbent telecommunications service providers, bursting of the “telecom bubble” and the drying up of necessary funding for BWA service providers and their systems suppliers.

Continuing progress - Nevertheless, fixed terrestrial BWA systems and services will continue making substantial contributions to the build-up of broadband subscriber access by complementing and competing with the wireline alternatives. The multiple implementation options listed in Section 1 provide the basis for these developments, and advantageous trade-offs lead to implementations that receive best market acceptance. The inherent wireless deployment flexibility makes possible “modular” service build-out that shortens time to market and makes possible gradual investment. These features facilitate end-to-end infrastructure ownership which enhances the service provider’s competitive position.

Commonalities and trade-offs - Existing and emerging fixed BWA implementations borrow from both conventional P-P microwave link technology and wireless mobile technology, but also introduce novel solutions. BWA operates in a number of frequency bands within a wide frequency range that overlaps with mobile service bands at the lower end, and extends to millimeter wave bands at the upper end. The basic commonalities of fixed BWA and mobile cellular wireless are in the usage of P-MP cells and in NLOS operation. However, different trade-offs apply because BWA services require substantially higher data rates, performance and availability, all of which reduce the potential NLOS service penetration in comparison with mobile cellular systems.

Innovation - BWA innovation is evident in several ways. For example, emerging MP-MP mesh networks are being introduced to substantially improve BWA service penetration and quality by establishing tandem LOS subscriber connections around obstacles. This makes it possible to serve more subscribers with better availability than in a P-MP cell using NLOS subscriber connections. In MP-MP mesh networks, each subscriber station simultaneously functions as a P-MP hub. Built-in alternative traffic routing, a distinctive capability of MP-MP systems, greatly improves subscriber access availability. This is a significant advantage over both P-MP and P-P systems which would require duplicating equipment to achieve a similar availability improvement. The increased hardware complexity of subscriber stations needed to achieve the greater MP-MP mesh network functionality is apparently realizable with readily available cost-effective technology.

Satisfying the different market demands - In both P-MP and MP-MP mesh networks the subscribers share the hub station capacity, which is advantageous where dynamic data rate allocation is used. However, subscribers requiring BWA service at higher data rates than can be provided by an optimized P-MP or MP-MP mesh network, will need to be served with millimeter-wave and/or optical P-P links. A redundant combination of mm-wave and optical links on the same path allows to substantially stretch the link length beyond the capability of either one to achieve the required availability. This is the most promising BWA approach to providing multi-Gb/s transmission rates to subscribers whose only other alternative would be a more costly fiber optic cable connection.

Outlook - The BWA outlook depends on technological and business prerequisites. This update leads to the conclusion that prerequisite cost-effective BWA technology is available and is satisfactorily advancing. The recent bursting of the “telecom bubble” clouded the business outlook. It thus remains to be seen to what extent BWA can be expected to satisfy the growing demand for broadband access at prices that end users will find worth paying for.

Acknowledgement

Harold Sobol, principal author of the TTG-3 Report [1] and Roger B. Marks, Chair, IEEE 802.16 Working Group on Broadband Wireless Access, kindly reviewed the successive drafts of this update and provided valuable suggestions for improvement. Additional editing help and was kindly provided by George Heiter, Past Chair of Committee MTT-16, Michael Green, Manager of Global Product Compliance, Atheros Communications,

and Peter A. Soltesz, President, PAS-COM. Thanks to all!

Reference:

1. Harold Sobol, Ferdo Ivanek, Sastri Kota, Amir Zaghloul, Bernard Bennington,

Clara Manders, David Weinreich and Eric J. Schimmel, Wireless Technologies

and Infrastructures for Subscriber Access, Technology Task Group 3, info,

Vol. 2, No. 2, April 2000, pp. 131-146.

................
................

In order to avoid copyright disputes, this page is only a partial summary.

Google Online Preview   Download