Modern travel means interminable waits, but it's a good time for reading. I finally read Wireless Pittsburgh: Sustainability of Possible Models for a Wireless Metropolitan-Area Network by Jon M. Peha, published in February as a working paper of the New America Foundation.
The good news — it’s full of interesting cost estimates and projected subscriber take rates based on specific demographics in Pittsburgh, Minneapolis and Philadelphia. The paper also examines a range of business models, in detail, from complete monopoly to structural separation (wholesale–retail) to let-the-market-take-care-of-it.
The bad news — all of the models turn out to be extremely sensitive to revenue assumptions, i.e. to the estimates of subscriber adoption and willingness to pay.
The flaws in this study
All of the models compute a net present value based on five years of stable operations, but there is no mention of technology evolution or adoption rates of competing broadband services, i.e. cable and telco (DSL or FiOS) services since this is a US study. Technology is evolving at a great rate. You can’t bet on stability.
During the past five years we saw WiFi go from 11 Mbps to widely deployed 54 Mbps systems and bleeding edge (pre-standard 802.11n gear) systems doing well over 100 Mbps. The last five years also saw costs decline to the point where we see widespread deployment of WiFi by individual consumers, a significant percentage of which are running open WiFi hotspots.
On a recent drive through three different neighborhoods in Portland Maine, I was interested in looking up real estate information on the web. On each of a half dozen occasions, I was able to find a open WiFi hotspot within one city block of deciding I wanted to connect. In January, I was in north New Hampshire and spent two nights at small motel (not part of any chain) in Littleton, NH. They had no Internet on offer, but a quick check for WiFi signals revealed two within range of our room. If you don’t like my anecdotal information, look at the WiFi hotspots that have been mapped by Navizon. It’s very different than five years ago.
No matter how much it simplifies the analysis, you can’t bet on stability.
What might happen with WiFi in the next five years? The latest WiFi specifications add multiple input, multiple output (MIMO) support, additional modulations and other goodies. As low cost WiFi routers incorporate these advances we’ll see speeds go over 300 Mbps, but more importantly MIMO technology increases both range and directionality. This means overlapping systems work better (despite their overlap) and the signals from isolated systems reach further.
If you’re worried today’s open systems will be locked down, then spend your time thinking about schemes like FON which offer more secure ways for consumers to share WiFi bandwidth.
If you’re worried consumer solutions won’t reach the inner city, perhaps someone needs to relook at where WiFi has already been deployed, and then forecast what might happen over the next five years, given the cost of a basic PC is approaching that of a television and the Cable and Telco duopolists both push triple play bundles.
Don’t short change technology evolution.