Private Networks For Public Safety

JonZittrain writes "Projects like the New American Foundation's Commotion are designing ad hoc mesh networking to keep communications open when governments want to censor. Former FCC Chairman Julius Genachowski and I argue that mutual-aid-based networks can be helpful for public safety, too, after attacks or natural disasters. There should be easy practices for anyone to open up an otherwise-closed Wi-Fi access point if it's still connected to broadband and is near people in trouble, and separately, to develop delay- and fault-tolerant fallback ad hoc networks so users' devices can communicate directly with one another and in a mesh. This can happen even while full packet-based ad hoc mesh is being figured out. The ideas have been developed a little in workshops at Harvard's Berkman Center and the FCC. Why not bring the human rights and public safety communities together towards a common goal?"

Comments

But... How will they listen in?

[earlzdotnet]

If we build a mesh network to communicate, then how will the NSA listen in? They'd have to dispatch someone to every disaster to ensure they had a node in the mesh that could listen in. That would cost us taxpayers way too much money

addressing and routing in adhoc mobile networks

[godrik]

I was always wondering how you achieve routing and addressing in a completely adhoc network. Let us assume that my whole city put their wireless access point in ad hoc mode even on the same SSID or whatever. How do you achieve any form of coherent addressing and routing ? You do not want to follow the ethernet technique with huge arp tables, devices are never going to be have enough memory to store such huge tables. Routing will be completely inneficient.
Any idea?

Re:addressing and routing in adhoc mobile networks

[Anonymous Coward]

B.A.T.M.A.N.

Re:addressing and routing in adhoc mobile networks

[Anonymous Coward]

You could reduce the ARP table down to two seperate, but much smaller tables: The first table is a list of known hosts and their last known routing neighbor, with aging and maximum size stack dumping. Keep this at about 1024 entries. The next table is a hash of the addressing schema which weights each of your local neighbors for each address range hash. You can keep this down to 65536 hash groups and highest scoring neighbor for discovered source address packets transmitted. You can expand this in one dimension by including instead of one neighbor, say 4 neighbors, in descending score on number of broadcasts in a given hash region. This could be done with IPV6 reasonably well. Even modestly powered modern ARM chips could easily handle punching through routing tables like those fast enough to keep packet delay down to reasonable levels.

If you're on the edge of the network, the hash table will likely only point to one or two neighbors that point inwards into the mesh. Inside of the mesh, you'll have a lot of excess packet passing due to the limits of the address hashing, but, packets will eventually find their way home, and each router will have a record of its most recent partners for getting back to the host.

Really, though, I don't see this working that well in anything less than 802.11N or AC. B and G just don't have the range or throughput in their normal usage (b devices had rather limited antennas, G devices may have range and speed boost, and slightly better antennas, depending on vendor, but, they still have limited throughput and range. Good quality N devices may have multiple radios and antennas, but are still fairly rare and non-trivially priced. The mesh will be a mess no matter how you look at it.

Ricochet did this post-9/11, routing worked fine.

[Myself]

While much of Manhattan's traditional communications infrastructure was literally a smoking crater after 9/11, the Ricochet mesh network was alive and well, built to barely notice the loss of individual nodes.

The company had recently gone bankrupt, but all the hardware was still in place, so some ex-employees drove from Denver to NYC with a bunch of modems and laptops, to bring mobile connectivity to the recovery effort.

Mesh works in this case because MCDN uses geographic routing -- the packet header literally contains a packed lat/long for the destination, and nodes make their routing decisions by angle and distance. There's a layer of name-to-geo resolution which makes that all work, and in the Ricochet days it was centralized, but I believe it could be made to operate with DHT like torrent networks do now.