Re-Thinking Mesh Requirements
12:35 PM -- Who was it that said less is more? Any debate regarding the origin and philosophical basis behind this pithy saying aside, such may indeed be true in some cases in metro-scale WiFi deployments -- and, if so, the cost benefits could be quite profound.
The general rule of thumb in mesh deployments is one backhaul link per three or four mesh nodes. Backhaul in this case refers to links between the mesh and external networks as opposed to the interconnection of the mesh nodes themselves, which is of course done over WiFi connections in most cases. WiMax is also sometimes applied here, but I don’t think that will last very long -- a single .11n link (.11n will shortly be appearing in meshes, and everywhere else) could easily swamp a single WiMax connection. An appropriate number of external connections via backhaul is critical, as we must avoid bottlenecks to throughput. And, of course, we want to avoid too many hops within the mesh so as to minimize latency, especially as the volume of time-bounded traffic increases.
But suppose we provision, via multiple WiFi radios per node, a lot more interconnection capacity within the mesh. This could have a very positive impact on latency, but it could also minimize the number of backhaul connections required in a given case for the very same reason. This could in turn have a similarly positive impact on the economics of metro-scale meshes, where backhaul is a major cost component. Sure, the backhaul connections provisioned would need to be faster if there are fewer of them, but the benefits could still be there.
You can read all about this idea, and a few other related items, in Farpoint Group’s latest White Paper on meshes, which you can find here. In the case of WiFi meshes, anyway, less really might be more.
— Craig Mathias is Principal Analyst at the Farpoint Group , an advisory firm specializing in wireless communications and mobile computing. Special to Unstrung