The Microwave Option

At a conference in San Francisco recently, I was touched with a bit of envy when I saw a colleague whip out a cable and connect her cellular phone to her laptop to check her e-mail. My jealousy increased when she launched a copy of Internet Explorer to show me some documentation she had posted to her corporate intranet. When she offered me the laptop I suddenly I realized my envy was completely misplaced. Only a user with the patience of a saint could put up with the slow speed she had to endure to surf the Internet.

Is wireless technology useless for the Internet? Can Internet access be delivered via wireless technology without users qualifying for Medicare in the meantime?

Having Internet data fly the friendly skies is an attractive possibility -- after all, there are already consumer offerings using satellites to deliver hundreds of thousands of bits per second to homes and businesses with a dish. Unfortunately, satellite schemes fail to deliver the multimegabit service of DSL and cable competitors. Is wireless broadband, therefore, just a fantasy?

There is one approach called Local Multipoint Distribution Service (LMDS) that has some possibilities. LMDS should have been a contender in the high-speed access marketplace. Unfortunately, that fact that you haven’t been offered LMDS service -- or perhaps even heard of LMDS -- is an indication of how likely it is to succeed.

LMDS uses microwaves to transmit voice, video and data in small cells between four and 10 miles in diameter. The technology is limited to small local cells because the low-powered microwaves lose strength quickly. But within those small cells LMDS has the potential to provide awesome amounts of bandwidth.

The possibility of avoiding the high cost of installing and leasing fiber means LMDS should be a winning access technology in corporate and university campuses. The cost of constructing an LMDS network is plummeting below $1,000 per subscriber, with most of the expense in customer equipment, which is seeing a natural decrease in cost. As prices continue falling there’s an opportunity for network and Internet access vendors to build high-speed access networks that provide better bandwidth at a lower cost than cable modems or DSL. At least that would have been the situation if vendors had reasonable access to LMDS frequencies.

Unfortunately the Federal Communications Commission (FCC) decided to take a different approach. In 1998 the FCC awarded a series of LMDS licenses to a variety of network vendors. It chopped each metropolitan market into three pieces: a large, full-service spectrum and a pair of tiny spectrums. Only the holder of the larger spectrum can deliver anything close to high-speed data service. As a result, the FCC effectively handed monopolies to the LMDS licensees in the larger sector.

Remarkably, the FCC also decreed that the LMDS licensee isn’t required to offer service with its spectrum during the first 10 years of its license. Also, the licensee never needs to commit its service to end customers. As an example, the LMDS license holder could simply sell the high-capacity local links it builds to local telephone companies. The telcos could use the wireless service to deliver traditional high-speed services and the LMDS license holder could simply pocket the savings.

This regulatory disaster is compounded by some fundamental technological challenges. First, no standards exist for LMDS. Just as we have suffered in cable and DSL Internet access, you’re forced to buy into your service provider’s preferred choice of hardware and networking technology. Second, LMDS delivers more than a gigabit per second over a microwave beam thinner than twine used to wrap recycled newspapers. That beam requires a "line-of-sight" between the networking carrier’s hub and the customer’s antenna.

LMDS microwave beams react with water, so over longer distances LMDS suffers from an effect called "rain fade." As the microwave travels through water -- from fog or a rainy day -- the signal power, and its ability to deliver high bandwidth, decreases. Because the frequency is so high, LMDS signals cannot penetrate walls, equipment closets or trees. That means simple landscaping can get in the way of the technology. Hewlett-Packard studied this effect in San Francisco and discovered that a networking company using a single LMDS hub would find 35 percent of its potential customers were behind objects that prevented the signal from getting to the customer.

Is there hope for LMDS? If you don’t start to hear about it soon, probably not. Still, there are a few encouraging signs. CellularVision in Brooklyn, New York, is using LMDS to deliver 1.5 Mbps connectivity for its Internet customers and doing it at $50 a month. An IEEE study group on wireless access is expected to deliver a plan for LMDS standardization.

LMDS is a potent wireless access technology that ought to compete with cable modems and DSL. Whether corporate or individual consumers of Internet access services ever get to see it is another question. --Mark McFadden is a consultant and is communications director for the Commercial Internet eXchange (Washington). Contact him at