The ManSig is trying to figure out how to build out and scale community networks. There are several key issues with wireless networks and community networks which are less of an issue with commercial wired networks. -- AdamShand
With community networks each node is often owned by a seperate person, further there is no paid admin staff to come to the rescue in case the network fails. Thus, it is important that the network be architected in a manner that faults are localized as much as possible. In other words the death of one node or router shouldn't be able to paralyze a large chunk of the network.
With wireless networks the key issue is that it is very geographic specific. Consumer grade wireless protocols (specifically 802.11b) run at high frequencies and have a fairly limited range (100 feet to a couple miles). This means that building a fully meshed network can be tricky through to impossible.
Because of these reasons we see AdHoc routing protocols as the key to our sucess in the ["NAN"] arena (the ["MAN"] will probably use a more traditional network structure).
See also: Peer2PeerRouting
- AODV (Adhoc On Demand Distance Vector Routing Protocol)
- DSR (Dynamic Source Routing)
- Grid Adhoc Mobile Routing Protocol
MadHoc (an AODV implementation)
MaNet (Mobile Adhoc Networks)
And some commercial software:
- Daman (Dynamic Access Mobile Ad-hoc Network ) created by a partnership between Alcatel and Sarnoff
FHP Wireless has a propriatry AdhocRouting program and sell Linux based AccessPoints.
- Mesh Networks was funded by DARPA and has developed a combined 802.11 and routing chip.