Metropolitan Wi-Fi Networks

There has been a great deal of press lately about municipalities building public Wi-Fi networks and opposition to those networks from commercial Broadband suppliers. Perhaps the most ambitious example of this is in Philadelphia. The stated goal for that project is to bridge the digital divide that now exists and prevents many individuals and families from obtaining the full measure of the opportunities generated by the Internet because they can’t afford the cost of wired broadband Internet access. While this is an admirable goal that I hope they are able to achieve, let’s take a close look at how they are planning to accomplish this.

My understanding is that Philadelphia plans on spending over $10 million to build a Wi-Fi (.11b variant) mesh network covering the city’s 135 square mile area and providing service to its 1.5 million residents. The estimated access point density is 8 – 16 per square mile. (or 1080 – 2160 total) This is based on the Wireless Philadelphia web site http://www.phila.gov/wireless/faqs.html. There wasn’t information regarding the yearly maintenance cost. 802.11b has a raw bit rate of 11 mbps and a much lower effective rate for the user. The bandwidth is also shared amongst all users of the access point. Further data rate reductions will be seen because of the mesh network configuration. Access points that don’t have a direct connection to the Internet will be required to forward packets from other access points. Depending on the frequency of access points with network connections, this adds significant overhead to delivering the data. In order to keep installation costs down and simplify the installation, it is stated that most access points will not have a direct connection to the Internet. Thus overhead will be high. If the typical definition of broadband is used, requiring at least 200 kbps down stream, this system runs out of bandwidth before the population can be served. On the other hand when a traditional broadband solution is deployed using DSL or cable, each residence can be delivered the desired data rate.

Taking all this into account, I don’t see how the stated goal can be met. I think the money spent on the initial build-out and yearly maintenance money would be better spent subsidizing those citizens that they are trying to help.