Posted by: Stephen Wildstrom on May 14, 2008
Earthlink’s announcement that it was throwing in the towel on an ambitious plan to bring public Wi-Fi to the entire city of Philadelphia marks the end of a glorious, but doomed experiment. The promoters of free municipal Wi-Fi hoped that optimistic techno-utopianism would somehow trump engineering and economic realities. They were, of course, wrong.
On an engineering level, the pro0moters of muni Wi-Fi got into trouble because they tended to underestimate badly the number of access points they would need to provide acceptable coverage. As the density of access points rose, so did the equipment budget, the complexity of the mesh networks, and the backhaul costs. The marginal economics of the original plans quickly became completely untenable.
Wi-Fi just isn’t a very good technology for this sort of wide-area coverage, especially in densely populated urban centers. It’s easy enough to provide coverage outdoors, where no one really wants it, but very hard for the signal to penetrate masonry or reinforced concrete buildings to get inside where people could use it. With access points located relatively close to the ground on lampposts or utility poles, It also proved difficult to provide service above the first couple of floors of even medium-rise buildings.
Earthlink says it will begin decommissioning the network in 30 days. In a statement, Greg Goldman, CEO of Wireless Philadelphia, the non-profit set up to run the service, says the organization and the city are "still working actively together to identify alternatives for preserving this network and applying it to numerous civic, commercial and social purposes." They may succeed in maintaining public hotspots in pockets of Philadelphia, but the dream of citywide coverage is dead.
Wide-area 3G wireless coverage, using a variety of technologies, is rapidly improving in U.S. cities, but there's little chance that it will be free. Verizon and Sprint operate extensive networks using Qualcomm's EV-DO technology, and AT&T's HSDPA network coverage is improving. T-Mobile is starting to light up its own HSDPA network.
Fourth-generation networks are waiting in the wings. With a deal struck between sprint and Clearwire and new funding in place, we should see the first serious deployment of WiMax technology next year. And sometime after 2010, Verizon and AT&T will roll out LTE networks using the 700 MHz spectrum they have just purchased. All of these approaches use licensed spectrum and require relatively expensive base stations and backhaul networks, so no one will be giving this service away. But there's a good chance it will be affordable, open, and a lot more reliable than public Wi-Fi.