The recent uprisings in Egypt and across the Middle East have caused some interesting echos amongst the great and the good back in Silicon Valley. Despite the overly dramatic language, I find Shervin Pishevar's ideas surrounding ad-hoc wireless mesh networks, embodied in the crowd-sourced OpenMesh Project, intriguing.
Back in the mid-years of the last decade I spent a lot of time attempting to predict the future. At the time I was convinced that personal area networks (PAN) were going to become fairly important. I got a few things right: that things were going to be all about ubiquity and geo-location, and that mobile data was the killer application for the then fairly new 3G cell networks. But I was off the mark to think that the convergence device, as we were calling them back then, was a dead end. Technology moved on and the convergence devices got better, thinner, lighter. Batteries lasted longer. Today we have iOS- and Android-based mobile handsets, and both platforms pretty much do everything my theorised PAN was supposed to do, and in a smaller box than the handset I was using in the mid-naughties.
It's debatable whether the OpenMesh Project is exactly the right solution to create secondary wireless networks to provide communications for protestors under repressive regimes. The off-the-shelf wireless networks Pishevar talks about extending operate on a known number of easily jammable frequencies. Worse yet, there has been little discussion, at least so far, about anonymity, identity, encryption and other security issues surrounding the plan. The potential for the same repressive regimes the protestors are trying to avoid being privy to their communications by the very nature of the ad-hoc network is a very real threat.
For these reasons I'm not absolutely convinced that mesh networking is the right technical approach to this very political, and human, problem. Perhaps he should be looking at previous work on delay-tolerant networking rather than a real-time mesh protocol.
However, Pishevar's suggested architecture does seem robust enough to cope with disaster response scenarios, where security concerns are less important, or if widely deployed to provide an alternative to more traditional data carriers. Pishevar isn't of course alone in thinking about these issues: the Serval Project uses cell phone swarms to enable mobile phones to make and receive calls without the conventional cell towers or satellites.
The Bluetooth standard proved too complicated, and at times too unreliable, a base upon which to build wireless PAN architectures. The new generation of ZigBee mesh devices is more transparent, at least at the user level, possibly offering a more robust mesh backbone for a heterogenous network of Wi-Fi, cell and other data connections.
With the web of things and the physical web now starting to emerge, and with wearables now staring to become less intrusive and slightly more mainstream, the idea of the personal area network might yet re-emerge, at least in a slightly different form. The resulting data clouds could very easily form the basis for a more ubiquitous mesh network.
I'd argue that I was only a bit early with my prediction. The technology wasn't quite there yet, but It doesn't mean it isn't coming. The 2000s proved, despite my confident predictions at the time, not to be the decade of the wireless-PAN. Perhaps its time is yet to come?