Friday, December 16, 2005

Will Municipal Wi-Fi Destroy the Value of Licensed Spectrum?
David Gross, Senior Telecom Consultant

Hardly a week passes now without a city or town somewhere in the U.S. declaring plans to create a local wireless network. From major cities like Philadelphia and San Francisco to suburbs like Aurora, Illinois and Plano, Texas, municipal governments across the country are designing and building new Wi-Fi nets primarily using mesh technology. The "mesh" simply means wireless access points are connected to one another, as opposed to just linking subscribers to the Internet. This makes it possible to cover large areas with reliable service, in spite of the wireless LAN protocol's limited range. Yet some license holders, most notably Verizon, are trying to stop people from using these networks, the underlying technology, or anything else that might encourage commercial use in unlicensed frequencies.

The rush to build mesh networks has less to do with mesh technology than it does cheap hardware. As a short-reach protocol, Wi-Fi has traditionally been viewed as unfit for outdoor applications. However, with 802.11 silicon reaching extremely low price levels, municipalities, not to mention college and corporate campuses, can deliver blanket coverage for less than it would typically cost with a much stronger LOS (Line of Sight) alternative like Free Space Optics, or even a NLOS (Non Line of Sight) technology like 802.16-2004, a.k.a. fixed WiMAX. Moreover, the mesh topology allows for re-routing around disabled access points, in a similar manner to how Internet routers circumvent network outages.

New Technologies, Old Lobbying Tactics

CDMA operators have responded to the surge of interest in 802.11 by offering expensive EV-DO (Evolution Data Only) service as an alternative, and by hiring lawyers to stop cities from building their own networks. This is an odd strategy considering that muni Wi-Fi projects are often intended to serve lower income households that cannot afford broadband. This consumer group is not a great target market for a service that requires a contract, costs up to $80 a month, needs a $179 PC card in order to work properly, and comes with a $35 activation fee. Considering the low cost alternative, EV-DO's growth would be better served by a price war than by a regulatory crusade.

In addition to being more expensive than Wi-Fi, EV-DO is much slower than the wireless LAN protocol. Its upload speed is barely faster than a dial-up connection. However, Qualcomm has started sampling chipsets based on the new Rev A standard, which will essentially match the transmission rates of a T-1. While this will do little to get EV-DO's prices down to Wi-Fi levels, it will make it fairly competitive with other traffic aggregation technologies, especially in urban areas where it can be difficult to find the large amounts of capital needed to build new wireline networks, and even more challenging to deal with the agonizing process of getting a construction permit.

Boring Backhaul

While it faces competition from other wireless technologies, particularly the recently approved mobile WiMAX standard, EV-DO Rev A will be an excellent way to connect urban Wi-Fi networks to the Internet. Muni meshes typically require at least one wide area uplink per square mile, and many of the light posts and rooftops on which antennae are placed do not have cheap or easy access to a wireline connection, which is a problem when the idea is to connect city residents to a global network. A long-reach wireless technology is the only alternative. But don't expect EV-DO vendors and providers to pump up this application. PR departments and the trade press would rather get their readers jazzed about streaming video on cell phones than go into the minutiae of aggregating traffic through a Wi-Fi to EV-DO gateway.

Remember that in the 90s, LMDS providers like Teligent and Winstar filled trade journals with exotic stories about videoconferencing through the air. Today, Teligent's spectrum is being used by First Avenue Networks to haul traffic for other wireless carriers. As EV-DO struggles to compete with Wi-Fi's low prices, it too has a promising future in traffic aggregation.

EV-DO's challenges are nothing new in this industry. With CDPD, GPRS, WAP, and MMDS, licensed wireless data services have an established history of being major busts. Nonetheless, consumers have increased their demand for cell phone minutes and science has limited the supply of radio frequencies. Spectrum owners should stop worrying about all the unlicensed antennae sitting on light poles, and start planning on connecting these networks to the Internet.