Metro Wi-Fi: RIP?

There's nobody I know who's done a better job of trackingthe metro Wi-Fi market than Glenn Fleishman, whose Wi-Fi Net News web site is agreat destination for people interested in wireless news analysis. That doesn'tmean I've always agreed with Glenn, especially at the intersection of wirelesstechnology and public policy.

Back in February, 2005, Glenn blogged about apoint-counterpoint article I wrote along with David Haskin, whose Mobile Pipelinewebsite was a venue for some of my writing. The topic was municipal broadbandand the role that metropolitan Wi-Fi networks were playing. At the time, I wasarguing that most metro Wi-Fi projects were ill-conceived and unlikely tosucceed. Because incumbent service providers were making similar arguments (butfor different reasons), I was accused by some of being a shill for entrenchedinterests. I even received e-mails suggesting that I was being paid by one ofthese companies to write my columns. Rest assured, the onlyfinancial transactions that have ever taken place between myself and a serviceprovider is when I pay my bill.

Unlike some analysts who opposed metro Wi-Fi based onarguments that it was unfair for government to compete with the private sector(the arguments were similar to those currently being waged about the publicoption in healthcare reform), my primary argument against these systems wasbased on the limitations of the technology and the lack of viable businessmodels. Wi-Fi is fundamentally a small-cell LAN technology. While it istechnically possible to build metro-scale Wi-Fi networks, it's not easy to do,requiring much greater cell density than traditional cellular systems oremerging 4G systems, especially if indoor coverage is a goal.

Greater celldensity translates into higher costs, both for build-out of the network andalso for backhaul services from the Wi-Fi mesh nodes to the Internet. And sinceWi-Fi uses unlicensed spectrum, it is virtually impossible to offer reasonableservice guarantees, especially in densely populated areas supporting thousandsof private Wi-Fi networks. The clincher lies in the fact that most peopleequate Wi-Fi with "free," so even if these companies were able to build out areliable network, people aren't willing to pay for it.

Glenn continues to write about metroWi-Fi companies abandoning the business and networks being shut down. Lastweek, he chronicled the adventures of Longmont Colorado, analyzing a recentarticle by Craig Settles, founder and president of successful.com and a long-timeproponent of metro Wi-Fi. Longmont found itself the owner of an orphaned metroWi-Fi network when its business partner Kite Networks went belly-up. NowLongmont has to decide what to do with it. I have a suggestion: cut your lossesand sell the equipment on eBay. Continuing to invest in this kind of network ishighly unlikely to benefit the citizens of Longmont. Truth be told, based onpersonal experience, I have my reservations about the ability of municipalgovernments to run customer-facing broadband networks, but even if the network civilservants of Longmont are special, it's still the wrong technology.The story gets really interesting when Comcast enters thefray, first by opposing a ballot measure that authorizes Longmont to act as abroadband service provider (it also owns a fiber network), and then byinstigating the establishment of a citizens' group established to oppose effortsby Longmont to leverage their network infrastructure. Both Settles andFleishman seem to enjoy vilifying incumbent service providers withSettles portraying Comcast as "The Empire" and their attorney as Darth Vader. Ican't help but feel some empathy for this point of view, given the abysmalrecord service providers have in greasing the wheels of government bybankrolling campaigns of politicians and mounting various secret andnot-so-secret lobbying efforts to advance their own interests. They providegreat case studies for the legalized corruption that is so typical of American publicpolicy deliberations.

Despite my skepticism about the wisdom of running a metroWi-Fi network, I'm hoping Longmont wins the freedom to make that decision. Thenotion that government has an unfair advantage in providing essential servicesto its citizens strikes me as capitalism taken to its worst extreme. Inparticular, the notion that governments should be prevented from building outessential network infrastructure (e.g., running fiber through establishedright-of-ways), reflects a warped view that government can do no right andbusiness can do no wrong.

If Longmont wants to run their ownmetro Wi-Fi network and their local political environment supports such astrategy, why should they be prevented from doing so? Is Comcast really afraidof the competition? In the end, it is likely that Longmont will conclude thatother civic priorities (education, public works, parks and recreation, etc.)are worthier endeavors with greater public benefits. Rapid technologyevolution and economies of scale argue against municipalities as wirelessservice providers. As for metro Wi-Fi, I think the future is clear. It's thewrong technology for the job.