We usually find ourselves recommending Wireless Networking in the Developing World to those that would like to learn about the technical aspects of what we do. For a compelling treatment of the political economy of networks try The Wealth of Networks. If you want a textbook treatment of networks in general, and you’re willing to pay, we recommend Computer Networking: A Top-Down Approach. If you’re not looking to pay, wikibooks has two relevant texts: Communications Networks and the narrower Computer Networks.
In general, We are building a network with no choke points – be they material or logical. It starts with the individual, and grows from there: the individual organizes the neighborhood, the neighborhood organizes the city, and so on and so forth, until you arrive at a global network that is owned and operated by every one and no one, for the good of us all. And yet the upper and lower layers of the networking stack need to be decentralized at the same time. Material decentralization is contingent on logical decentralization. That is why projects like tent.io, Freedombox, sneer, et cetera are so important. If the logical flows remain centralized, we will not be able to take advantage of the efficiencies of local routing, not mention still being spied on by purveyors of ‘false p2p’ (G+ and FB being prime examples). This is not going to happen overnight: it will be gradual, and from the inside. It is already happening, and we view its continuation and evolution as inevitable.
FreedomLink and FreedomTower are network appliances. FreedomLink is designed to anchor a regional network coop, and FreedomTower to anchor a neighborhood one. The regional network would be made up of neighborhood networks, linked together. Mostly, these projects consist of the integration of existing tools into a tested and turnkey suite. The other project of note is the FreedomNode, which would anchor a home/business/building network. One of the key ideas of the whole project is that we can implement transparent crypto on the node, so that all traffic is encrypted end-to-end.
It’s all in the details. There are limitations to mobile ad-hoc – this is what most folks understand “mesh” to mean: that the nodes can move around. Free networks are much bigger than mesh, though – they can be implement in fiber, copper, fixed wireless, or mobile ad-hoc. It is a matter much more of political economy than of technology. So, the solution is simply not to try to make mobile ad-hoc scale. Isolate link-dynamic media access domains into pockets of managable size, and join those pockets together with more static links. In this way you can scale, still offer a mobile experience, and use whatever technology is most appropriate for the geography.
DNS is something of a separate question, but we see no reason why a distributed naming scheme shouldn’t come to prominence in the next few years. There are several in the works. The general idea here is to use a DHT for discovery of names that are self-authenticating and globally unique, but ugly (like BTC addresses), then, once you’ve discovered the machine-readable name for the resource in question, simply give it a pet name, like we used to do on AIM.
There are certainly challenges ahead, but our vision is well within the realm of technical feasibility. There are already sizable cooperative networks in Spain, Athens, Berlin, Kabul, Nairobi and a host of other cities. These last-mile networks can be connected securely using tunneling while research continues into low-cost, long-haul communications platforms.