Strange times in Portland, Oregon: The Unwired Portland project is aimed at delivering Wi-Fi access to every part of the city, but some people just don't get it -- the signal, that is.
According to a recent story on CNN's Web site, the project engineering left something to be desired. The contractor engineered the system -- and no doubt, the bill that Portland has to pay -- with some 25 Wi-Fi access points every square mile. This, it seems, would place all users within 500 feet of an access point. Unless you want to sit on your front porch, however, or perhaps be willing to walk down the street a bit, you might not get any signal at all.
Hmmm. Some users might resolve the problem by purchasing and installing an outdoor antenna or bridge. Although the network operator notes this is a US$150 cost, a quick look into its online store shows that some antennas would run a user close to $300 including installation. Some might not consider that to be "free wireless."
The alternative, according to the provider, would be to quadruple the deployment of access points so no user would be more than 250 feet away from a signal source. The leader of the project sums it up nicely in the CNN story: "The network cost gets completely out of whack. The business model breaks in its entirety."
So, what went wrong? Two possibilities jump to mind, both unsettling. The first: We assume the network provider has integrity but lacks the ability to engineer a metropolitan network appropriately. Not good. The second: We assume the provider knows how to engineer a metro network properly but, as the project leader noted, a design that quadruples the number of installed access points probably would never have gotten through the proposal stage.
Perhaps the goal of the network provider was simply to win the bid (with a low bid, most likely) and worry about the effectiveness later. (This is pure conjecture because I was not party to any of this.)
What I do know is that not all Wi-Fi products are created equal. We've benchmarked many products and have seen differences in device rates and ranges that conservatively can be called dramatic. In the case of Portland, the difference between deploying one product or another could be the difference between success and failure. One test we published last year using 802.11a (rather than 802.11b/g) showed a single client could maintain 18Mbps throughput at 500 feet -- the magic number for Portland. Additional tests showed throughput of more than 10Mbps and 1,500 feet -- and even 1Mbps at 1,700 feet.
Thus, equipment matters. Nonetheless, a scan of the network provider's Web site provides no readily available information about the technology being used to deliver their metro Wi-Fi coverage. Truth be told, this is no real surprise. Today's metro Wi-Fi providers seem to be following the practice of the ISPs: They are more than happy to tell you how great their service is, but fail to reveal anything about the infrastructure that they are putting in place to deliver that service.
Let Portland serve as a lesson to anyone building metro or campus LANs. Before you sign that purchase order, make the service provider sign a service-level agreement guaranteeing you'll get what you paid for -- without having to pay more.