 |
|
|
The goal of broadband fixed wireless manufacturers and carriers is to provide an alternative to wired high speed internet connections for consumers and businesses. Technology proponents hope to squeeze into the market currently dominated by cable modem and DSL services. To do this, a mixture of radio technologies and concepts will have to be used, drawing from different parts of the radio industry. Broadband wireless could prove to be not only a viable competitor in urban areas that currently have high speed internet options, but could carve out a new market niche in the many places in rural Canada that are not wired for cable and that are too far from Telco switches to obtain DSL.
However, the market outlook for broadband wireless is not altogether rosy. There have been many recent failures of broadband wireless carriers, and without a major breakthrough in sales, many of the vendors are likely to go out of business, or get swallowed up by a larger competitor. Greater market acceptance is required, and standardization may play a significant role. There are even proposals in the US to use the LMDS 2.5 GHz spectrum for 3G wireless, which could be the kiss of death for this entire industry.
Broadband wireless is a product of the convergence of different forms of wireless communications. Microwave has long been used for high bandwidth, network backbone applications, such as connecting a cellular base station to an MSC. Microwave is restricted by line of sight requirements (the two antennas must be pointed at each other) and is used to communicate between two points. It is therefore not suitable for connecting many users together. It is a backbone technology, not an access technology.
Initial concepts for LMDS (Local Multipoint Distribution Service) and MMDS (Multi-terminal Multipoint Distribution Service) were for distribution of content from a base station to multiple terminals, using small antennas mounted on the outside of the homes, apartments or offices of customers. Line of sight would still be a requirement, but one base station could now service multiple customers, bringing the cost down dramatically, although perhaps not far enough.
At the other end of the spectrum is cellular or PCS communications. This has always been point-to-multipoint, but has in the past been oriented to relatively low speed voice communications (8–13 kbps). In order to support mobile devices, line of sight restrictions had to be eliminated. Signals can bounce off multiple sources and still recombine to provide a usable communications channel. These advances gave cellular and PCS quite good in-building penetration.
Combining these ideas results in the ideal broadband wireless system:
|
Frequency of operation of radio devices is gradually becoming less of a technical issue, and more of a legal, regulatory and political issue. Technology can adapt to an ever wider range of frequencies, but they have to be available for use over a wide geographical area, and allow the desired technology to be used. Two frequency bands that are set aside for broadband wireless in North America are LMDS (Local Multipoint Distribution Service) in the 27.5–31.3 GHz frequency band and MMDS (Multi-channel Multipoint Distribution Service) in the 2.1–2.7 GHz band. European systems are more likely to operate in the 3.5 GHz band.
Some companies are choosing instead to avoid licensed spectrum and are heading for the 2.4–2.5 GHz ISM (Industry, Scientific, Medical) band, or other bands which do not require the purchase of spectrum or coordination of frequencies. However, some observers are concerned that this band will rapidly become saturated by a multitude of uses and eventually become unusable by everyone. While an unlicensed band makes it cheaper and easier to get a service up and running, it may not be the best choice for the long run.
Many vendors offer equipment that works in multiple bands, so that a migration from an unlicensed band to a licensed band is a possibility if a service becomes a victim of its own success, and more assured access to spectrum becomes a requirement.
There are a number of solutions being offered that provide the first steps in wireless broadband – point-to-multipoint communications, with small antenna footprints suitable for installation on the outside of a house or office. These systems are generally based on a single carrier modulation scheme, although some use frequency hopping or time division to try to reduce the risk of putting all your eggs (data bits) in one basket (narrow slice of frequency). The requirement for external antennas to maintain line of sight communications on most of these systems becomes a major cost disadvantage.
One of the biggest limiting factors for widespread implementation of broadband wireless as a competitor to wired high-speed internet connections is price. It is generally agreed that US$300 is the barrier that vendors have to squeeze below to become competitive with wired internet services. And, this is the installed price, including in-building wiring and labour costs! Most vendors have not yet achieved this. Standardization will be critical to increasing volumes which, in turn, will lead to the lower prices that can make the whole industry viable. It is only natural, however, that every vendor with a proprietary technology wants it to become the single standard. However, technical issues also have a significant impact on costs.
Removal of line-of-sight requirements is crucial to the success of broadband wireless. Traditional microwave involves communications between two points, and nothing but air must be between them. Broadband wireless systems based on this also require line of sight between the base station and the terminal’s antenna. While this is not an insurmountable problem for point-to-point microwave connections, it will increase the number of base stations that are required for broadband wireless, and probably require outside antennas for consumers, dramatically increasing costs. Furthermore, if the location of the base station is critical to maximize the number of terminals that can be covered, the cost of the site may well go up, as there will be fewer locations to choose from, and the owners of the sites will know this.
If broadband wireless systems cannot penetrate buildings well, externally mounted antennas are required. This significantly increases the labour and equipment costs of installations, due to the wiring and the installation and orientation of the antenna. Eliminating this restriction is probably necessary to reduce costs low enough to crack the consumer market.
The Broadband Wireless technology space is suffering more from a shortage rather than a surfeit of technology choices. Vendors are banding together, hoping to be part of the winning solution, if a standard is chosen.
It might appear that a standard is superfluous, because broadband wireless is largely a local service. As long as the base stations and home or office terminals are compatible in each area, what is the advantage of a standard?
The most direct benefit is reduced costs through higher volumes of equipment, and the entry of companies with products that would not be cost effective if they had to be customized for each proprietary protocol. More effective test equipment, for example, might be a result of a widely accepted industry standard. Companies are unlikely to customize equipment for 20 different protocols, resulting in the availability of only the most generic link analysis capabilities.
It is worth remembering that, without the standardization of AMPS, the cellular market in North America would not have achieved its current size, at least not until GSM, the second major cellular standard, came along.
Many of the benefits of standardization cannot be predicted ahead of time. It is unlikely that anyone would have predicted that cellular phones would ever have been used as mobile data modems, yet that happened when standardization resulted in increased sales of voice terminals, more R&D and eventually units small, robust and cheap enough for data applications.
If there ever is a widely accepted standard for broadband wireless, it will probably be based on OFDM (Orthogonal Frequency Division Multiplexing). It can break through the line-of-sight restrictions that plague older, single-carrier solutions. This radio divides a stream of data into multiple parallel parts, carrying each at a different frequency at different times. With diversity in both time and frequency, it is quite robust in the presence of noise, interference and multi-path fading that occurs when signals bounce off multiple objects.
The Wideband OFDM (W-OFDM) camp is led by Wi-Lan of Calgary, which has patents on many aspects of this technology. Wi-Lan is one of the founders of the OFDM Forum (www.ofdm-forum.com), which has a number of major players behind it, including Alcatel, Ericsson, Motorola and Nokia. Wi-Lan’s W-OFDM, known as Multi-code Direct Sequence Spread Spectrum (MC-DSSS) spreads the information being transmitted over a wider bandwidth than standard OFDM, allowing individual carriers to be spread farther apart, and therefore reduces the effects of frequency errors. Wi-Lan claims that their technology is also more resistant to multi-path fading and fading of portions of the frequency range.
Wideband OFDM is not only being considered for use in broadband wireless, but it is also used by ETSI BRAN HiperLAN/2 and the 802.11a standard, which is a recently developed, higher performance second generation of the popular 802.11b ‘wireless ethernet’ standard known as WiFi.
On the other side of the Broadband Battlefield is the Broadband Wireless Internet Forum (BWIF; www.bwif.org), led by Cisco, but also boasting Texas Instruments, Analog Devices and Andrew Corp. in their army. Vector OFDM (VOFDM) uses two antennas to add space diversity to the time and frequency diversity already provided by basic OFDM. Since fading of signals is often very different at two places just a short distance apart, the use of two antennas significantly increases the possibility that a usable signal will be received at the other end of the channel.
In the middle of the pitched battle, attempting to mediate a settlement, rather than allow the war over broadband standards to continue, is the Institute of Electrical and Electronics Engineer’s (IEEE) 802.16.3 standards subcommittee. They would like to repeat their success with 802.11b, an indoor wireless ethernet replacement standard that is far ahead of HomeRF and Bluetooth in acceptance and implementation for small scale indoor networking.
According to Roger Marks, Technical Contact for 802.16, the IEEE standards organization has already approved a single carrier solution (i.e. not OFDM) in the 10–66 GHz frequency bands for publication as 802.16.2. It is expected to be publicly available before the end of 2001.
OFDM proponents are happier about a second standard, known as 802.16a, which will provide for both single carrier and OFDM approaches in the unlicensed 2–11 GHz bands. Keeping OFDM proponents even happier will be a third standard, 802.16b, which targets the 5–6 GHz band and will only support OFDM. Both of these are scheduled for availability in 2002.
The Wideband OFDM camp was well represented at 802.16 meetings, and provided a great deal of input. The BWIF forum, however, made no official representations. A third group, known as the DSL Consortium, was so happy with the work that 802.16 was doing, that they closed up shop entirely. This organization, including Nortel, decided that their goal of ensuring an industry standard was developed had been attained.
There is some confusion over the status of the BWIF at the IEEE, because although it did not participate in 802.16 standardization, it is a program of IEEE-ISTO (Industry Standards and Technology Organization). However, this IEEE affiliate organization merely provides management services for BWIF and is not a participant in IEEE standardization, despite its name.
It appears that the Wideband OFDM camp may have won the standards battle, but the war is not over yet. Many a standard has been published and then languished in obscurity. The real battle is always in the marketplace. Standards have considerable clout, but then so do large corporations such as Cisco.
ComDev is one company that has taken an interesting approach to broadband, choosing to use high speed CDMA cellular/PCS protocols instead of the single carrier or OFDM approaches. With CDMA aiming for higher and higher data bandwidths, and with huge production volumes of cellular/PCS terminals and network infrastructure, their strategy is aimed at producing low costs and small form factors. This may particularly appeal to rural carriers that have many customers unable to access the internet via cable or DSL, and much unused spectrum to give over to high speed data services. One of these carriers is Alpine PCS in rural California, which is hoping to find a new revenue stream by being able to offer voice services and high speed data on their CDMA systems.
This path also brings up the issue of roaming. At first sight, roaming with a high speed internet line does not seem to make much sense. But, there are probably many businessmen who regret being downgraded from Megabit to Kilobit speeds when travelling, or being forced to pay high prices in their hotel room to get their access back again. If broadband wireless was provided over cellular or PCS protocols, and if the high speed wireless data ‘modem’ was small and portable (perhaps even a PCMCIA card), customers could simply pack it along with their laptop and have high speed internet access anywhere, anytime.
This approach does require that adequate spectrum be available for a large number of high speed data users, and be widely implemented. If this approach proves to be less applicable in dense urban environments where idle spectrum is simply not available, roaming may still not be practical.
Broadband wireless is perhaps a case of too many cooks spoiling the broth. But, after a period of consolidation of carriers, vendors and technologies, a small number of broadband wireless offerings may survive to provide a new, more flexible way for Canadians to access the internet at high speed. It is certain that if this technology gains a foothold, and the questions about the industry’s financial viability become a thing of the past, many applications for internet access will become practical that are simply unthinkable with tethered access.
© – Copyright