Hacker Newsnew | past | comments | ask | show | jobs | submitlogin

"If it wasn't for those damned laws of physics always screwing up my plans..." basically. Mesh networks have a pretty hard constraint of k/n^2 bandwidth total where k is an average node's bandwidth and n is the number of nodes. You quickly get down into the 'bytes per second' range.


This interests me, but I can't find a paper - stuff like "mesh network bandwidth" simply don't work on scholar. Any chance of a citation?


Jinyang Li, Charles Blake, Douglas S. J. De Couto, Hu Imm Lee, and Robert Morris: Capacity of Ad Hoc Wireless Networks. Mobicom 2001. http://pdos.csail.mit.edu/papers/grid:mobicom01/

Summary: Throughput of a mesh asymptotically approaches 1/7th of the non-mesh link rate. This is using the 802.11 MAC, but recent softMAC chipsets could be run in a more mesh-optimized psuedo-802.11 protocol (that has not been invented yet AFIAK), so you might get performance as good as 1/4th.


1/7 or 1/4 is a lot better than k/n^2.


I think this may be because no one has attempted it on a scale where it is essentially cripplingly slow.




Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: