The market for broadband wireless access services is expected to grow dramatically in the next several years, and vendors are scrambling to get systems ready for the emerging market.
Fixed terrestrial wireless systems are expected to generate global service revenues of nearly $10 billion during the next five years and $28 billion in 10 years, according to a new report by The Strategis Group.
“Global interest in wireless broadband is really taking off with potential in both developed and developing telecom markets,” said Susan Welsh de Grimaldo, senior consultant at Strategis.
Vendors agree the market is being driven by a bottleneck in the last mile. Several companies have been laying fiber, but at around $250,000 per mile, fiber is not a cost-effective solution for any but the largest of businesses.
A solution has been lacking for small- and medium-sized business. Wireless broadband access using point-to-multipoint systems appears to be at least one of the most promising ways to bridge the last mile between the high-speed fiber networks and businesses.
How the voice and data traffic will be sent over the airwaves, however, is still up for discussion in what many believe could become a debate that will rival the one between Code Division Multiple Access and Time Division Multiple Access technologies in mobile systems.
Two air interfaces exist for transmitting traffic in point-to-multipoint systems. The established model is Frequency Division Duplexing, which divides the available spectrum into two separate channels-one for transmitting traffic and one for receiving traffic with a guard band in between. A newer approach called Time Division Duplexing, allows traffic to flow in either direction on the same channel, but in different time slots.
FDD vendors have a clear lead because the technology has been used for years and is proven. Netro Corp., which has developed an FDD-based point-to-multipoint bandwidth-on-demand product called AirStar, already has commercially deployed its system. Earlier this month, Netro also announced an agreement with Lucent Technologies Inc. under which Lucent will distribute Netro’s AirStar product under a private label.
Although TDD technology has been used in the Personal Handyphone System in Japan and Digital European Cordless Telecommunications systems in Europe, it has yet to be commercially deployed in a broadband wireless access system. TDD backers believe, however, TDD technology will be the second-generation of wireless technology for broadband systems.
Netro said one of the advantages of a system like FDD, which has been around for a while, is that it can make use of existing components rather than investing in developing new components. In addition, Netro said it can be difficult to make the transition from a trial system to a commercially viable system transmitting traffic, which is something Netro has already had to accomplish.
FDD backers also point out that most of the spectrum allocations around the world are designed with FDD in mind with two separate frequency allocations and a guard band. TDD backers contend that wouldn’t prevent their systems from being used in those spectrum allocations.
Stuart Feeney, vice president of engineering at Netro, said FDD provides for a continuous and simultaneous dialogue between the base station and the subscriber because transmissions are not broken up into time slots. A continuous downstream to all subscribers, he said, allows for scheduling of information to subscribers.
At the heart of the debate is the need for networks that can handle data traffic without compromising voice quality.
Proponents of TDD believe that their system is the best answer to the increases in data traffic. Data, they say, behaves differently than voice, which is almost always synchronous. Instead, data traffic is asynchronous and bursty.
With data traffic, a customer might need a big chunk of down link to access a Web page, while only requiring a small chunk of up link to navigate the Web. The situation would be reversed if the customer sent a large document, for example.
The key, they say, is that you can never be sure exactly what the requirements will be, and TDD allows the flexibility to maximize the efficiency of the system. TDD is often compared to a highway in which traffic is moving in both directions, but the median can be moved back and forth in near-real time to allow large amounts of traffic to get through in either direction.
“Because you can burst in both directions, you can get an average of 30-percent to 40-percent greater throughput,” said Thomas van Overbeek, chief executive officer of WavTrace Inc., which recently signed an agreement to install a beta test system for Formus Communications Inc. “The more bursty the traffic the better the benefit of TDD.”
San Diego-based Ensemble Communications Inc. also has developed a system based on TDD technology that includes three adaptive layers.
Netro, however, said at the end of the day, traffic often averages out to be symmetrical, a phenomenon it says is increasing with applications like e-commerce in which more and more traffic is transmitted back to the base station.
In addition, Netro said FDD systems create their own efficiencies. By allowing transmit and receive to occur simultaneously, transmissions can occur at the same rate as TDD but only use half the channel, the company said.
Carlton O’Neal, vice president of marketing and sales at Ensemble, agrees that TDD systems do require some separation between transmit and receive called a guard time, which is similar to guard bands in FDD systems. Guard times do create some channel loss, he said, but less than what is lost with FDD systems.
The U.S. Local Multipoint Distribution Services B-block creates an interesting challenge and opportunity for vendors. TDD proponents believe TDD is the only system that will work in that block because the guard band is theoretically smaller than it needs to be to avoid interference in FDD systems.
Netro’s Feeney concedes the LMDS B-block provides one of the more difficult challenges for FDD systems, but he said although the guard band is smaller than ideal, it is still sufficiently large enough to operate an FDD system.
“It might be technically challenging, but compared with the worldwide access problem, one frequency block is minor,” he said.
There are drawbacks to TDD, including the need for very precise synchronization, which limits the distance information can be sent within TDD systems. Ensembles’s O’Neal, however, said it is unlikely wireless broadband access systems will be called on to transmit more than a few miles.
