Who knows whether mesh networks will become the third broadband option, but there is no denying that in times of disaster, mesh networks add up to a technology that simply works. The New York Times looks at the role of mesh networking in the current Katrina disaster recovery efforts; and how the mesh-based WiFi networks were used to jerry-rig a communications infrastructure in the New Orleans region.
WiFi meshes elegantly dodge our phone system’s central problems. They’re low-power and ultracheap – and decentralized like the Internet itself, which was initially conceived to withstand a nuclear attack. You can use WiFi to build a do-it-yourself phone system that is highly resistant to disaster.
The Times, thankfully does point out that despite failing in high-stress situations like New Orleans, the centrally switched phone systems works more often than not. Even the cellphone networks, that are overloaded during emergencies, can handle traffic, though not as optimally as the wired phones. Still, it makes absolute sense to develop a battery-powered mesh-based wifi system , that acts as an emergency communications network.
In the case of an emergency communication network, it doesn’t seem financially practical to me. While I’m by no means an expert on the subject, I was always under my impression that a significant amount of nodes need to be deployed for a mesh network to be sustainable.
Attending IEEE’s Radio and Wireless conference in Boston last year, mesh networks were a focus for researchers and pioneers.
Somewhat unrelated, I remember researchers mentioning the social aspect and fairness was a major concern.
Placing inexpensive “nodes” in items we see everyday, for instance cars or light bulbs, by moving around, users are handed off from node to node (or if someone is stationary, a passing car temporarily carries the load before handing it off to another passing car).
In urban locations, such as New York City, the load per node would be spread around evenly. But, if a person should happen to be on the fringe or more rural areas where there are fewer people to share the load, many users along the outskirts will use one node, causing significantly more drain and “unfair.”
Allen, the idea of mesh for emergencies is not that you have a permanent network with high density but that you bring in mobile nodes to handle the load during the emergency.
Om, a GSM based mobile network can be very good in times of emergency because GSM supports identity based call prioritization (i.e. you can set the system up to kick users off to make room for emergency response users, as needed). At one point around the turn of the milenium GSM careers in the US were soliciting federal contracts to provide emergency services based on this feature, but Sprint and Verizon made a stink about how that would be unfair to them and that GSM was a filthy “French” technology.