Whither Wireless?

Published in IEEE Spectrum Magazine, Sept 2013

A recent report of PCAST (President's Council of Advisors on Science and Technology) recommends spectrum policies that would lead to greater capacity for wireless communications.  The cellular bands are nearly filled, new spectrum is hard to come by, and the modulation techniques are already optimized for efficient spectrum utilization.  What's left to do?  The report concludes that the answer is to move to smaller and smaller cells.  As the cell size shrinks, the capacity becomes virtually limitless.

I imagine a world filled with wireless picocells.  But it seems to me that we already have something approaching that with the spread of WiFi hotspots.  As usual, I'm writing this column in a coffee shop and taking advantage of their free public WiFi.  Their little cell just covers the dozen or customers surrounding me now.  There's another cell next door, and then at the door after that, and so on all down the street.  In fact, when I drive the streets of this little community, I'm never out of range of a WiFi hotspot.

The web site WiGLE lists about 2 billion sites that have been discovered by wardriving contributors, and the total number of WiFi networks in the world is estimated to be about 6 billion. When I zoom into the map of hotspot locations I see that in my little town there is a dense concentration of sites along all the main roads, and I know that the unmapped side roads are also dense with sites.  My own house, however, is not on a road, and an ordinary computer cannot receive the signal from my home network from a road.  Nevertheless, to my constant amazement, Google Maps knows exactly where I am when I use my iPad at home.  Just how do they do this?

Google says that they use crowd-sourced WiFi hotspot data to triangulate the location of networks.  I assume that the only way they could locate me would be to combine information from networks reported by several neighbors before reaching a known street location.  However they do it, I'm impressed, and it says something to me about the density of WiFi coverage and geographic usability of this data.
In spite of this density of coverage, when I walk out the door of this cafe, I will have to rely on my cell phone and the services of my cellular provider.  Needless to say, this is expensive -- especially in comparison to the free coverage inside the cafe.  But almost all those other networks outside the door here are closed to me.  Many are encrypted, and others are pay sites.  WiGLE reports that only about 17% of their listed sites are unencrypted, and that this percentage has constantly decreased with time.  I'm reminded of the old saying "water, water everywhere, but not a drop to drink."  The picocells are everywhere; I just can't use them.

I have this imaginary scenario where I convince everyone in my town to open up their WiFi networks.  We all use Skype for telephony, get rid of our cell phones, and get cheap, pre-paid cell phones for travel and emergency.  Of course, there would be a worry about privacy with open networks, though privacy could be maintained with encryption and other safeguards while leaving the network access open. Moreover, I think many people don't give this a second thought when using public WiFi today or when posting their lives to social networks.
I wonder how this scenario would evolve.  Maybe open networks would spread throughout the country and the cellular providers would lose a lot of business.  Investment in the cellular infrastructure might disappear.  WiFi locations might get overloaded and experience interference problems.  Even today some experts are predicting the imminent collapse of WiFi,  but maybe, as in the old predictions of the collapse of the Internet, WiFi will continue to thrive.  In any event, what role it plays remains to be seen.  Right now WiFi is seen as an adjunct to the cellular network, but it is conceivable that this paradigm could be reversed.  Conceivable, but I'm not placing any bets on this.