Part 1: Three Things Google Needs To Understand About Broadband Wireless Internet Access
I have no idea how accurately articles like Google Goes Wireless in last week's Business Week Magazine accurately reflect what Google really thinks... or is conflicted... or confused about "helping" to stimulate Broadband Wireless Internet Access.
But if the articles do accurately reflect Google's thoughts or intent... In my opinion, Google is going way, way off the mark, trying to "muscle" the telecom industry through activities such as lobbying the FCC and partnering with companies such as EarthLink. On that path lies ruin because Google is trying to compete on the well-worn playing field of the telecom industry incumbents, who, while weakened, are by no means incapable of fighting off newcomers like Google. Not to mention partnering with companies such as EarthLink who have only entered the Broadband Wireless Internet Access business out of desperation, not out of strength.
Having observed the Broadband Wireless Internet Access industry nearly from its birth in the early 1990's through the present, and observing from a completely independent perspective, I have some ideas about how Google could effectively accelerate real competitive Broadband Internet Access nationwide through the deployment of Broadband Wireless Internet Access. What I will propose could be done from Google's real strengths, rather than from its weaknesses.
1) Not one type of Broadband Wireless Internet Access... but many forms of Broadband Wireless Internet Access
In trying to promote Broadband Wireless Internet Access, the first thing that Google has to understand is that there are many forms of Broadband Wireless Internet Access... not just one, or even a few. There are many different Broadband Wireless Internet Access technologies, deployment methodologies, and business models. Merely "bringing more 700 MHz spectrum" onto the market will do almost nothing to accelerate competitive Broadband Internet Access deployments. If that were the case, we would have already seen the effect - one set of 700 MHz auctions has already happened, with large geographic footprints purchased by entities other than telecommunications industry incumbents... and we're still hoping for something to be different... better... with the upcoming 700 MHz auctions?!?!?!
Much has been made about Google's experiments with its Google built/owned/operated Metropolitan Wi-Fi network in Mountain View, California. I've been in the audience at a presentation where Google representatives have told stories about the hard lessons they have learned from the Mountain View network, leaving the distinct impression "well... we know we're never going to be doing that again!" But I'm convinced that the majority of Google's bad experiences were the result of Google simply not knowing much about real-world deployments of Metropolitan Wi-Fi networks, too much faith in the blanket recommendations of their vendors, and the use of first-generation, inadequate-for-the-conditions Metropolitan Wi-Fi systems.
In between the potential "reach" of 700 MHz systems, and the short range of Metropolitan Wi-Fi systems lies a plethora of systems whose capabilities extend to miles... as well as tens of miles. There are more than enough Broadband Wireless Internet Access technologies, deployment methodologies, and business models to "bridge the gap" of truly competitive Broadband Wireless Internet Access. One example, from personal experience, is how CONXX, Inc. has commercialized a rural "fixed wireless" infrastructure network called AllCoNet2 into a system specifically designed to address the severe challenges of telecommunications (not just Internet) infrastructure in rural counties. As a result of AllCoNet2, there is no longer a "digital inclusion" challenge in Allegany County, MD because AllCoNet2 (which CONXX operates). CONXX effectively combines multiple Broadband Wireless technologies and systems, including point-to-point microwave, proprietary (pre-WiMAX, odious as that term is to me) Broadband Wireless systems, and second-generation Metropolitan Wi-Fi systems.
With technology such as that offered by Meraki, and the wireless technology developed by and embedded in (and hopefully commercialized at some point) in One Laptop Per Child, it's possible for individual groups of ordinary Internet users to form their own networks to provide Broadband Internet Access.
2) In Broadband Internet Access, Smaller, More Decentralized Is Often Better
Google can be forgiven for not understanding this point because their corporate "DNA" was formed at the core of the Internet, where it has best been served by providers of massive scale - fiber bandwidth, huge data centers, massive amounts of electrical power, etc. But the challenge of Broadband Internet Access is to bridge the last mile(s), and often the "best" way to do so is very specific to the challenges of a particular area. In the case of Allegany County, MD cited above, it simply wasn't cost-effective to deploy a county-wide fiber network and so Broadband Wireless technology was used. In other areas, fiber-to-the-premises is indeed cost-effective. Very often, small, entrepreneurial Broadband Internet Service Providers are in the best position to bridge the "Broadband Last Mile". Sometimes, municipal governments need to step in. Other times, very small telephony companies, electrical utility companies, rural electric co-ops, even farmer's co-ops serve as catalysts to provide Broadband Internet Access. Google needs to understand this dichotomy between the realities of providing services such as its "Googleplex" at the Internet core, where massive scale is absolutely required, and the last mile(s), where the scale (can be) much smaller and thus more effective.
3) In The US, Licensed Spectrum Is Always Gamed
In the US, because Congress has mandated that new spectrum allocated for commercial use be auctioned, an entire industry has been created because there are enormous profits to be realized from "winning" spectrum at auction. No matter how tightly-crafted Congress, the FCC, and independent experts think that spectrum auction rules have been crafted, US spectrum allocation by auction will, simply, always be gamed (artificially manipulated) with results other than what was originally intended. By its nature, licensed spectrum that is exclusively allocated will favor incumbents. That's because an incumbent is "in the best position to "make effective use of" additional spectrum or has ample incentive (and financial clout) to purchase additional spectrum to "warehouse" it and thus reduce potential competition. Consider that there are, depending on how one counts "blocks" of spectrum, fourteen or more individual "blocks" of spectrum, each of which could support a company that provides Broadband Wireless Internet Access service:
- 700 MHz auction (~ 2005)
- 700 MHz auction (2007)
- Two blocks for analog wireless telephony @ 800 MHz
- Six blocks (or nine; some areas had a block subdivided) for digital wireless telephony @ 1.9 GHz
- One block for digital wireless telephony @ 1.9 GHz for Sprint / Nextel for its "800 MHz swap" to resolve interference issues with public safety communications
- 1.7 / 2.1 GHz band auctioned in 2006
- 2.3 GHz band
- 2.5 GHz band
All of these blocks of spectrum "exist" in rural areas as well as urban areas... yet Broadband Internet Service Providers that are actually operating in and providing service in rural areas only rarely have access these blocks of licensed spectrum because auction rules overwhelmingly favor well-funded incumbents who can outbid small service providers and then "warehouse" the spectrum until they deign to use that spectrum to "provide services" (always comfortably off in the distant future) or to simply thwart the rise of potential competitors.
In short, Google needs to recognize that Broadband Internet Access is every bit badly in need of reform and reinvention as the Web Search Engine business was several years ago... and even more potentially lucrative, on a number of levels. Broadband Wireless Internet Access using license-exempt spectrum is a core precept that should be as important to Google as open source software (Linux) and open protocol networks (Internet, TCP/IP) to its continued success.
I'll conclude this article later this week in Part 2: What Google Should Do To Promote Broadband Wireless Internet Access.
By Steve Stroh
This article is Copyright © 2007 by Steve Stroh.
(Continued apologies to readers for light postings recently as I continue to recover from a major system crash.)
Thanks Steve for an article that hits the issue right on the head.
Its so nice to hear someone talking sense from experience!
People who should know better a clueless about what it takes to deploy wireless broadband. They think they have valid "common sense" views of how wireless propagates, but all their real experience are:
Radio and TV with their one way, hndreds of Mhz spectrum and megawatt transmit power, huge guardbands and super simple recievers (Works great in my house, oh wait I have cable..)
Cellular Phone with low Ghz spectrum,4kbps bandwidths and pretty small cell sizes (Usually works ok in my house)
802.11 with pretty big channel bandwidth but very short range and almost purely line of sight. (It works great in my house)
Posted by: Robert Berger | August 12, 2007 at 13:31