A couple of weeks ago there was this whole fracas over Net Neutrality: Google reversing its stance and teaming up with Verizon to treat the wireless spectrum differently than the wired infrastructure, and now AT&T suggesting the same thing. It’s a sticky subject that’s attracted even the nutzo Tea Party crowd to the fray. In the end, though, does any of this even matter? Whereas achieving good coverage with wired infrastructure is fairly expensive, and once deployed becomes a fixed asset, achieving the same coverage in wireless spectrum need be neither. All that’s needed is sufficient manpower and the right combination of technologies, some of which we already have.
I’ve posited before (or at least hinted here and here) that growth in wireless ad hoc networks could represent a threat to existing centralized service providers. Further, we could start creating and deploying these right now with technology we already have. Any wifi or otherwise wireless capable device can be repurposed to serve as either a router or an access point in an ad hoc network. Achieve sufficient density, which depends on terrain, weather, and other obstructions and interference sources, and one could blanket an entire city with not especially robust or impressive wifi access. Improvements in hardware and software, especially ad hoc routing, along with better compression will improve transmission throughput and signal strength considerably.
Where do we get these wifi devices? Landfills are a good place to start, followed by all those cell phone and electronics recycling outfits. First off, it’s better if they don’t end up in the landfill, and second, people will gladly give them away (at least until there is some value associated with them). It’s not enough to simply have a processor, although these are available in greater abundance. The devices have to have some wireless capability: wifi, 3G/4G radio, bluetooth.
Once we have the devices, the next step is to repurpose them. This could involve either firmware flashing using a custom firmware hacked for each device, or pulling the transmitters out of the original devices and throwing them into custom kit, then using a much smaller number of tweaked firmwares. The level of effort for each approach may be similar, though at first glance the second option is appealing. In the case of 3G/4G radios, it may be sufficient to create custom SIM chips that handle the network access, but in all likelihood each device will need a special operating system loaded. The result should be what amounts to a custom built microcell like the ones AT&T is selling to its customers to make up for poor reception. The only difference is that this one is not licensed, making it technically illegal.
Next up is the battery life. This is, in fact, the most significant limitation of current technology. We simply don’t have batteries that are capable of operating at the power level we need for this application and the durations necessary, and it’s unlikely that anyone would want the job of going around to replace batteries on dead microcells. That means we need to have other ways of maintaining power. Combinations of wind and solar are good for this, but there may be times when both are simultaneously inadequate. In those times, portions of the network will likely degrade. New technologies (such as capturing static electricity from the air) may prove beneficial for this application.
Finally, to round out the hardware, all of this needs to be weatherized and ruggedized. It is unnecessary for anyone to access the screen and keyboard of recycled mobile phones, so these can be sealed up. Access to the battery compartment is still advisable, however, in case batteries need replacing. Then again, if these devices are cheap enough to make, then simply deploy a new one when old ones die. In any event, there is no reason to subject the devices to the ravages of the elements more than necessary, and the devices should be quite functional even after such weatherization.
Assuming we’ve been able to flash a bunch of devices with a custom ad hoc routing firmware, the next step is mass deployment. Unlike the biomimetic ad hoc routers I’ve mentioned previously, these are entirely reliant on the willingness of people to “lose” them in strategic places, preferably out of sight and not easily reachable. Sometimes this means climbing to rooftops. The questions of where to deploy and what density is sufficient depends entirely on the self-healing abilities of the mobile ad hoc routing protocols in play, the ratios of different wireless technologies (blutooth has a shorter range than wifi, and wifi has a shorter range than 3G and 4G, for instance), local terrain, the locations of buildings, weather patterns, and the locations of electromagnetic interference sources. Because the resulting microcells have a much lower transmission power than existing cell towers, the average coverage area of each microcell will be far smaller than the roughly 10 square miles covered by existing cells. Further, there is a possibility that frequency overlap might result in the cell towers overpowering the microcells (of course, given enough density, the microcells could be powerful enough, being closer, to overpower the cell towers, but more investigation is warranted here).
OK, so this is a good time to talk about the legal aspects of such an operation. That is, this whole thing would likely be illegal, in the same way it’s currently against the law to cause wireless interference on certain (most) frequencies right now. With the exception of walkie talkies and CB radio, personal transmission over radio spectrum requires approval by the FCC. Some frequencies are reserved solely for use of various entities, while others are doled out as necessary or as requested by licensed operators. It might be possible to transmit through current wifi frequency spaces: the 2.4, 3.6 and 5 GHz bands. Since transmission through these bands (especially the 2.4 GHz band) probably requires FCC licensing, you could expect authorities to actively hunt down and destroy transmitters they find. This is even more certain if you add on encryption, which will be necessary to secure the ad hoc network against unwanted intrusion. The long term viability of such an effort will require dedication, probably grass-roots organization, and lots and lots of hardware.
Can it be done?
GooglePlus
Twitter