Network Computingwww.networkcomputing.com

Evaluation

Overall, LMDS compares favorably with competing options on both a performance basis and a cost basis, but it lacks the wide support and financial backing that other platforms possess. This could prove to be significant.

Industry support has flooded behind cable modem and ADSL technologies. Low-cost chipsets and ASICs are available for platforms, and the technologies boast an impressive list of supporters, including Lucent Technologies (Murray Hill, NJ), 3Com (Santa Clara, CA), Analog Devices (Norwood, MA), Compaq (Houston, TX), Intel (Santa Clara, CA), Microsoft (Redmond, WA), Motorola (Schaumburg, IL), Ericsson (Sweden), NEC (Japan), and so on. Members of the computing industry in particular seem to be supporting ADSL, and to a lesser extent cable modems, as a means of delivering multimedia content to homes and businesses (a vehicle to justify the purchase of faster processors, greater storage capacity, and new software). This multibillion dollar industry has a lot to gain from the success of broadband delivery, and logic suggests that its members will go to great lengths to bring the most probable broadband technologies to the mass market. ADSL seems to be that technology.

LMDS is not as fortunate. Only a few companies have publicly committed to supporting the platform, and those which have, with the exception of HP, lack the distribution, name-brand awareness, and financing that supporters of ASDL and cable modems possess. The difference is likely to result in LMDS CPE that cost more than and lacks the distribution of cable and ADSL modems. In addition, there will be lower visibility for LMDS. Time also could be an issue. If cable modem and ADSL services become widely accessible within the next 2 years, then deployment of LMDS could prove unattractive in areas that already possess other alternatives. The cost per actual subscriber of LMDS infrastructure would shoot to over $3000 if penetration falls below 17 percent, and high levels of penetration could be tough to come by if 38-GHz radio, ADSL, fiberoptic links, satellite systems, and cable modems are already in place.

Thus, what will happen to LMDS? Only time can really tell. There are a number of variables that could drastically alter the market and the fortunes of LMDS providers. ADSL and cable modem deployment could lag considerably behind expectations, and satellite and many LMDS operators may not even build out their networks.

However, developments in the market today suggest that leading LMDS auction winners will deploy networks and that these service providers will concentrate their efforts on business and well-to-do residential customers. The high cost of CPE will preclude deployment to other residential areas, at least initially.

LMDS will face stiff competition from established and incipient technologies. And this is anticipated to cause LMDS to fall considerably short of the FCC's vision of it as a universal high-powered competitor to existing cable operators and ILECs.

It is unlikely that LMDS will come close to the FCC's mandate for significant coverage of operating areas in the short term, and this decree could prove to be unfeasible even in the long term. Yet, despite this, LMDS operators could be very successful in their own right. It is not LMDS's ability to succeed that really needs to be questioned but rather its ability to live up to the hype that surrounds it. In all, the technology's greatest shortcoming could prove to be the unrealistic expectations that have been cast on it.

Competing Technologies

One of the most promising great equalizers in the access network is wireless local loop technology, also known as broadband wireless access (BWA). Think about it: New competitive carriers can deploy wireless networks in a fraction of the time it takes to build wired infrastructures. And the availability limitations of DSL and cable modem services are not an issue in wireless broadband networks. BWA holds the promise of getting high-speed, competitively priced services to enterprise sites, remote offices, and, well, just about anybody really fast.

Two primary technologies are being rolled out to enable BWA. The first has been discussed in this chapter quite extensively: local multipoint distribution services (LMDS), a line-of-sight technology running in the 28-GHz band. LMDS is most suited for densely populated urban areas, where it is difficult and expensive to deploy additional or new wired infrastructures. Typical speeds are 45 Mbps downstream in a point-to-multipoint configuration. However, LMDS has the potential to exceed OC-3 (155 Mbps) speeds. Distances between sites are limited to 4 km.

The second primary technology—multichannel multipoint distribution service (MMDS)—operates in the 2- to 3-GHz band, is less susceptible to interference than LMDS, and has no line-of-sight requirements. MMDS can support greater distances than LMDS—up to 30 miles between sites. The tradeoff is that MMDS is slower, delivering downstream speeds in the neighborhood of 10 Mbps.

With both technologies, an antenna and radio are installed on the roof of a business site and are connected by coaxial cable to customer premises equipment in the LAN wiring closet. Also, commercial service availability is imminent. WorldCom has been conducting MMDS service trials with schools and residential and business customers in Boston, Dallas, Jackson, Mississippi, Baton Rouge, and, Memphis using equipment from Cisco and Motorola. Meanwhile, Cisco has said it plans to begin commercially shipping LMDS and MMDS interfaces for its routers by midyear—which sounds like any day now. The move could prove to be a major stepping stone for service provider deployments.