Our series takes a look at Chapter 7: Local Multipoint Distribution Service (LMDS) Design Technology.

As mentioned previously, local multipoint distribution services (LMDS) auctions were completed on March 25, 1998. The 128 rounds of bidding resulted in sales of 864 licenses for a sum of $578,663,029.

According to the rules set by the FCC, cable TV (CATV) operators and ILECs were not allowed to bid on A licenses in their territory, and their participation was light. US West was the only ILEC to bid, acquiring eight B licenses for $10 million. Auctions raised considerably less than the $3 billion that was expected, causing some parties to label it a failure and dampening enthusiasm on Wall Street.

However, the limited amount of revenues that was earned may be more of an indication of the size of the bidders than of the actual potential of the technology. It also may be evidence that bidders learned a lesson from previous personal communications services (PCS) auctions.

With LMDS auctions finally behind them, license winners and industry watchers now need to focus on a new set of problems and unknowns. LMDS proponents are not alone in their quest to bring wireless broadband and bundled services to businesses and consumers, and the fate of the spectrum that has been awarded is up in the air. Auction winners may opt to wait 3 years until they can sell these licenses to local exchange carriers (LECs) or cable providers. Questions also linger regarding the financing of many license winners and their ability to establish an LMDS network within a short period of time. Time to market could become very important because there are several other options that could fill the same void that LMDS carriers seek to address within the next 2 years. Background LMDS is essentially a fixed cellular network that operates at millimeter-wave frequencies. The system overcomes the challenge of signal attenuation, which can be a significant factor at millimeter-wave frequencies, by limiting the distance that signals must travel. Cell sizes in LMDS systems are on the order of 2 to 5 km, whereas cell sites in analog cellular systems are spaced 2 to 3 miles apart in metropolitan areas. As a result, reliable service can be obtained with relatively low output power. Communications are two way, and systems will be digital for the most part in order to maximize bandwidth efficiency and service quality. Perhaps the most intriguing aspect of LMDS is its wide bandwidth and capability. A-block license holders will have a 1150-MHz slice in the 28- and 31-GHz bands, which is large enough to support CATV, local telephony, and wireless broadband access simultaneously to users in each cell.

The cellular architecture also will enable providers to tailor CATV programming and advertising on a cell-by-cell base if desired. Yet, despite its many capabilities, the technology's success is likely to depend on provision of broadband and bundled services around data rather than voice or CATV. Why?

The latter market is already overcrowded with several alternatives, including traditional cable operators, whereas competitive local telephone service in itself is unlikely to justify the cost of LMDS customer premises equipment (CPE) for residential customers and many businesses. The ILECs' record for reliable service also will be a difficult obstacle to overcome in the telephone market. LMDS telephony is likely to be used for additional telephone capacity and sold to existing users.

Data, on the other hand, offers LMDS providers a golden opportunity. Bandwidth-hungry Internet and intranet content is pushing the telecommunications backbone to its limits and hampering the effectiveness of existing corporate and consumer data connections. Integrated services digital network (ISDN) basic rate interface (BRI) service is proving to be inadequate, and T1 services are too expensive for most users. Meanwhile, digital subscriber line (xDSL), cable modems, and satellite broadband programs still lack low-cost CPE and wide deployment. LMDS service providers could obtain a significant share of the burgeoning demand for low-cost, high-bandwidth data services in the business market if they can deploy a reliable network around major telecommunications users in a timely fashion. However, this may prove to be more difficult than originally anticipated. Stalls in FCC auctions have given 38-GHz radio, cable modem, and xDSL providers a head start over LMDS operators, and many questions remain about the ability of LMDS providers to build up a network quickly. As a result, LMDS will face some level of competition. An important question, then, is how does LMDS fare against competing options.

Competitors

Asymmetric digital subscriber line (ADSL), the xDSL front runner, will be able to provide from 1.5 to 7 Mbps downstream and from 200 kbps to 1 Mbps upstream of data to customers using the existing telephone company's twisted pair copper wires. Approximately 60 to 90 percent of U.S. access lines will be able to receive ADSL service with little or no upgrade to existing infrastructure, but ADSL solutions still cost from $900 to $2200 per line, significantly more than consumers or ILECs care to pay. Competing standards (CAP) and discrete multitone (DMT) have injected uncertainty in the market and slowed its progression, as has lack of interoperability between modems made by different manufacturers. Resolution of these problems may be forthcoming, however. Low-cost application-specific integrated circuits (ASICs) from leading semiconductor houses promise declines in the price of second- and third-generation ASDL modems. In addition, the Universal ADSL Working Group (UAWG), which is backed by a who's who of the networking, telecommunications, computing, and semiconductor industries, is working toward resolving lack of interoperability. The UAWG's work, if successful, also should help bring about lower-cost modems.

The other leading technology in the race to bring broadband capacity to users, cable modem service, can provide data rates of up to 36 Mbps downstream. However, because the medium is shared, experts consider speeds of 10 Mbps downstream and 200 kbps to 2 Mbps upstream (for two-way modems) more realistic. Cost per home passed is about $700 to $800, and interoperable multimedia cable network system (MCNS)–compliant modems are expected to be widely available for $400 to $500 by year end 2001. Two-way service is available to approximately 20 million U.S. households and businesses today, but support for the service has waxed and waned in recent years, depending on the provider. Some major cable providers appear to be more interested in digital CATV and telephony rather than in upgrading infrastructure to be two-way cable-modem-capable. Regardless, the technology is available today, and it works. Expan sion is expected in coverage as well, although it is unlikely that more than 60 percent of U.S. homes will be passed by two-way cable-modem-ready infrastructure within the next 5 years.

In addition to ADSL and cable modem providers, satellite system operators and visionaries such as Kirkland, Washington–based Teledesic are now offering two-way data services to users around the world. Published expected data rates of the various satellite constellations vary between 200 kbps and 2 Mbps downstream/upstream for residential users and between 10 and 30 Mbps for corporate users. Production of CPE is limited to a few players per constellation, and online dates are in the year 2002 at best. Satellite systems are best suited for providing access to areas that are sparsely populated and which lack other alternatives. The cost per home passed for satellite systems can be below $30, but CPE exceeds $1700 in cost.

The other land-based alternative, fiberoptic cable to the home or curb, is prohibitively expensive and is deployed only to heavy business users. Full installation of fiber in the U.S. local loop would require at least 20 to 25 years to be completed at full speed, and the ILECs have done little to suggest that they are in fact pursuing fiber to the home or curb at full speed. Competitive local exchange carriers (CLECs), including long-distance giants, and the ILECs continue to establish fiber in densely populated areas, and these systems are already operational. However, extension of this capability beyond heavy business users is unlikely.

Despite a pipeline of several hundred megabits or even gigabits to each user, the cost of fiber deployment to the home (well over $3000 per home passed) will preclude widespread installation for some time. Sometimes referred to as a fiber pipeline in the sky, 38-GHz radio is operating successfully in several major metropolitan areas. To date, it has been used primarily to provide other telecommunications carriers with additional capacity. However, a shift to point-to-multipoint networks and the resulting use of low-cost CPE (less than $4000) should translate into a rapid customer expansion. Most important, 38-GHz radio is operational today.