TechDirt has some scathing analysis of Chaska, Minn.: Remember a few months ago when I complained that Chaska was the poster child of metro-scale Wi-Fi? That none of the metro-scale equipment vendors and deployment firms could point me to urban environments with large-scale working networks? It turns out, the Chicago Tribune reports and TechDirt dissects, that Chaska's network didn't work well at all until recently. I had heard weeks ago that a Chaska official winked at a conference when asked about the cost of replacing all its early Tropos gear--the Tribune says it "shelled out more money" for the newer transceivers.
The Tribune paints a fairly bleak picture and quotes Bradley Mayer--then Chaska's network go-to guy, now part of EarthLink's efforts--as naive about RF propagation. "Mayer made some unpleasant discoveries. Like the fact that wet leafy trees absorb radio signals, hampering Wi-Fi coverage. And this one: Wi-Fi signals don't pass through stucco like they did wooden walls, another negative for coverage."
Since these are obvious and well-known issues with Wi-Fi and RF, I'm hoping Mayer was misquoted. They shouldn't have to learn that; it was well known by the time I started this blog in 2001.
TechDirt says it best when it notes, "Of course, the Tribune article claims that that [network troubles at Chaska] is all in the past, and now the network is great. Well, I'm sorry, but there is a credibility gap to address now. We were told it was 'great' 18 months ago, but later we're told by the same person 'It took about a year and a half before we felt we really had a good handle on the network.' "
I feel a bit taken in, too.
This is why there's a necessity for independent testing and verification of claims and performance that should be written in to every single RFP that's being issued, for the benefit of the companies installing the networks, the cities authorizing or encouraging them, and the population that will use them. No company should be able to show its own numbers. Independent RF and performance audits should be mandatory, and city officials should have to publish the results.
We've moved a long way away from the issue of taxpayer-funded networks, but that doesn't mean networks can't perform, and that those in charge can mischaracterize them. As we're seeing in Mountain View, Tempe, St. Cloud, and elsewhere--the first networks with many thousand of simultaneous users--it's not as easy as any one in the industry has been pretending.
I've touched on this a bit on my blog as well - wide scale wi-fi is really, really hard. Especially in dense areas, as you need many more access points in a much smaller area to cover many more people. That's a recipe for massive interference and confusion. Throw all the existing AP's into the mix and you've got lots of problems. It wouldn't be so bad if we were using 5 mhz channels, but with 20 mhz channels, it's insane.
Using WiFi LAN technology for wide area mesh deployment is the wrong solution. Even a rudimentary radio engineering analysis would have told them they where violating radio physics and good radio engineering principles. When are these WiFi mesh equipment companies going to start doing more RF engineering and less marketing hyperbole? I foresee that even larger scale future installations of this WiFi mesh equipment will be metro scale disasters.
I am a resident of Chaska and subscribe to the Wi-Fi. We're thrilled at the low price of $17/month compared to the $45/month thru Sprint. We had some great luck with Chaska's service awhile ago, but in the last 1-3 months, coverage has been spotty. We live in a stucco home next to mature trees and it's been raining a lot. Go figure...Seems like we pick up the neighbor's service (not chaska.net). I'm not technical in nature, so all I can tell you is that the solid orange light is on a lot on the router and that's not good.