Fluid networking

During my travels through the Amazon interior, I visited a forgotten river town called Murú, in the state of Pará. There are thousands of towns like it, all of which suffer financial and infrastructure deprivation and neglect from the state government.

There is no cellphone service in Murú, much less internet. To communicate with the outside world, one has two options: take a bus to the next city, or climb about 300 feet up the microwave backhaul tower, which undoubtedly cost more to construct than the rest of the town combined. This tower doesn´t serve Murú at all, it is simply an outpost that relays traffic to more important places.

Upon talking to the leader of the town, who is city hall (the prefeitura), I learned that three years ago, they did have internet, which served only the school. But since then, a the then-newly-elected politician in the region pulled the plug on it, and moved to the capital, apparently also laundering away whatever infrastructure budget there was designated to improve the school, roads, waste management and communications infrastructure. This is normal practice in the north of Brazil.

The town effectively doesn´t exist in the state government´s burocracy. The denizens are on their own. The only economic activity there is the importation of prepackaged junk food, snacks, soda, and beer. The plastic detritus of these mom-and-pop enterprises never leaves, and simply accumulates.

It is unlikely that this town or others like it, will receive any assistance from the outside world to upgrade any of its infrastructure; as it is, in spite of the natural beauty of the place, it is unfit to grow a tourism industry. It is the third world part of Brazil.

I thought many days about what could be done to help places like this pull themselves up by their own bootstraps. One idea was to implement a WikiOrganization Town Hall template to help such towns with administration, since nobody else is going to do it for them.

But first, they need internet! They don´t have the financing to install high-tech microwave or fiber, or even DSL. What to do?

Fluid Network

• Android cell phones as a freely-flowing link to the wired world
• There is no reason an open-source phone OS has to connect only to cell phone towers, traversing the cellular network. The system should be able to be modified to function in peer-to-peer mode, where trafic traverses along the roads and rivers populated by other cell phones in p2p mode.

The mesh networking is a great solution that is affordable for municipalities, but not poor villages. A main problem in the interior of the Amazon is that there are often several hundred miles between cities or towns that have cellular service, much less 3G or another internet provider. In otherwords, there are no access points on the backhaul connecting two regions. One remote city I am familiar with (Tucuruí) has no backhaul connection to the internet fiber at all, everything is routed through a single satellite link that is horribly overloaded, and suffers the latency problems of satellite.

The roads connecting localites that do have service are dotted with quaint villages (some of which are little more than lemonade stands). Exceptionally well-off villages might have a church and a school. For places like these, there is no fincancial incentive for the wireless companies to build out infrastructure.

My idea of a fluid network consisting of p2p cell phones is to at least create a wireless bridge to the nearest cell phone tower that has data service, whereever there is sufficient traffic composed of volunteer phones. Generally, these roads always have a stream of trucks, buses, and cars. If an Open Source phone OS like Android become common in an area, it could create a virtual network, and even offer a bit of offline storage for message queueing (using the onboard flash storage). It wouldn't be fast, but it could handle basic communications (like email and sms) and perhaps serve small schools with a usable link to the internet.

I think this could be done simply by turning the business model of the android inside-out, so to speak, where the mobile nodes seek out neighboring phones to pass traffic and stored messages, when a route becomes available to a celular data service.

The problem here is who will pay for the equipment and who will pay for the connectivity. Ideally a publicly owned mesh, or a mesh composed of volunteer nodes (something like the HAM operators of the old days) could solve the problem of paying for infrastructure and bandwidth.

See also

• Freifunk (Free Radio)
• MORE protocol - MAC independent Opportunistic Routing, an opportunistic routing protocol designed for wireless mesh networks
• Commotion Wireless
• TORouter - Torouter is the codename for a hardware project to provide an easy to setup Tor bridge or relay. It could also act as an optional wireless access point with all traffic transparently routed through Tor. The end goal is to have an easy to use system that is a bridge or relay by default, functions as your Internet router, and is around $120 or another hopefully reasonable price point.
• Breaking bottlenecks - a new algorithm enables much faster dissemination of information through self-organizing networks with a few scattered choke points