In the US, cities such as Philadelphia have opted to use Wi-Fi technology to fill in broadband Internet access gaps, where the incumbent carriers and ISPs are not providing broadband coverage. The moves have been politically interesting: they sparked a legal row resulting in Pennsylvania passing a law that gives the right to veto any public-sector Wi-Fi (read the archives at Wi-Fi Networking News for more details).

That law may be a particularly US-centric phenomenon, but there is also discussion on the technical feasibility of city-wide "hotzones", in which big-name pundits have pooh-poohed the efforts.

Their reasoning:

  1. Wi-Fi was developed as a LAN, intended to cover a span of only about 300 feet
  2. There's no proven economic business model for Wi-Fi in the metro-area or WAN, because no "terrestrial data-only system" has ever made money
  3. 802.11-based networks run in unlicensed spectrum, and as interference increases, these networks could become unreliable.

There is some merit to all these points. But there are also strong arguments against them. Let's look at each.

  1. Wi-Fi is a LAN technology?
    It's true that, wherever possible, it's preferable to deploy the right tool for the job at hand. But where there's a pressing need, you generally use the tools available, rather than simply not doing the job at all - particularly when those tools are inexpensive, mature, standard, and well proven. Not everyone can build a cable, DSL or metro fibre network; but nearly anyone can build a Wi-Fi network. Maybe they don't reach very far: they're cheap, so you just put in more access points to extend coverage and capacity.
  2. There's no money in it?
    If cities are deploying Wi-Fi networks in an effort to get real rich, real quick, it's true they might be disappointed. But perhaps their goals are loftier. Many cities, economically, feel the need to attract and retain denizens, in part, by competing through modernisation of their infrastructures. And while it's true that no mobile wireless data-only service (emphasis on the "mobile") has been a rousing success, wired data services (Internet access services, as a humongous example, not to mention frame relay), certainly have been. And that's what these Wi-Fi services are. As in the case of Philly, they are Internet access services in places that currently can't get them via DSL, cable or some other medium because the usual suspects haven't offered them. Whether access is over the air or via a wire is, frankly, irrelevant.
  3. Interference will kill them?
    The unlicensed spectrum issue might be a legitimate worry, a fleeting issue or a non-issue. First of all, a number of self-governing, channel-hopping, single-channel and power-adjusting systems are emerging to minimise interference. Second, the addition of 802.11a networks to the mix in the 5 GHz band gives us many more non-overlapping channels to work with: 11 to 24 depending on geographic location. Finally, WiMax (IEEE 802.16), the much-ballyhooed technology for metro and WAN wireless last-mile coverage, which barely registers a pulse on the reality charts, operates in both unlicensed and licensed bands. And which do you think are the most popular with carriers looking at early deployments? That's right - the unlicensed bands, to keep deployment costs down.

In the interim, we have Wi-Fi.

Might Wi-Fi broadband service eventually fade into the sunset, usurped by more appropriate technologies as they emerge and their costs fall? Sure. But who cares? Wi-Fi is dirt cheap, and its usefulness sends a fast payback.

So is there any particular reason not to solve current access needs with it while we wait around for WiMax or for incumbents to roll out broadband services using other platforms?

Absolutely not.