State of the Art? A 'Pringles® Cantenna'

Meraki is a well-funded Silicon Valley startup that sees a future in disrupting the walled-garden world of proprietary networks; the private cable, copper, fiber or licensed spectrum as marketed by ATT, Verizon, Comcast and the like.

Meraki (who have raised north of $20M so far, including a visible chunk from Google, where recently departed WiFi and spectrum muckymuck Chris Sacca acted as midwife to the investment) is presently in the news for their plan to give away some 15,000 WiFi repeaters to folks in Baghdad-by-the-Bay. If this succeeds, they think, the dreamy landscape of free citywide WiFi access will come to pass, first in San Francisco, and eventually, all over the place.

Although Henry Blodget's daughter Kelsey (I presume that is the relationship) has a pretty positive post about this plan over at alleyinsider.com, I beg, on scientific and social-network grounds, to disagree with some of her assumptions.

Meraki has nabbed most of MIT's Roofnet team. Roofnet (which didn't work in Cambridge) has top-flight people -- I have the privilege of knowing some of them -- but the real problem is this; WiFi itself just doesn't work very well! It's OK for short range, line-of-sight applications. However, WiFi frequencies don't penetrate obstacles well or go around corners, and they're readily absorbed by water molecules -- quite common in leaves. Think I'm ranting? Read some Claude Shannon (the god of information and communications theory) and decide for yourself.

Add the shortcomings of the underlying technology to the simple fact that the network itself will not provide seamless roaming access unless fifteen thousand or so folks ante up their own personal cash to build this out, and you are heading toward yet another WiFi user experience disappointment-fest. Like any network model (see, passim, Reed's law about the value of social networks), it takes a visibly successful (ie, useable) initial rollout to attract more rollout nodes - success will feed on success, as it does in (for instance) the Twitter world, but failure also feeds on failure. And unlike the social nets built on software, the Meraki model (essentially a hardware social net) ain't free to join.

That said, there are things right around the corner for open access to spectrum which likely will work. They're based on 802.16 (WiMax) and not 802.11 (of any flavor.) The mobile mesh packet handoff stuff that came out of MITRE's labs works well regardless of the underlying transport frequency mix.

In short, Blodget has part of the story correct - mobile meshes are the future. But she's betting the wrong horse if she thinks WiFi will make it to the finish line first - or at all, IMHO. There will NEVER be a successful nationwide Muni WiFi rollout. The best will in the world won't make 802.11 radio waves behave in a reliable, scalable manner. And I will bet that as soon as viable and inexpensive WiMax gear hits the market, Meraki will transition over to stuff that actually works.

I write about this for Broadcasting and Cable as well, most recently here where I look at the announced entrants in the 700MHz spectrum auction. I'll post a longer version of that article here on MediaZulu soon.