Extended Medium Observation (EMO) — A load-aware metric for routing in wireless mesh backbone networks

The main advantage of using specific routing algorithms like AODV-ST or OLSR in wireless mesh backbone networks is their optimized handling of throughput from border nodes to each other. However, currently used metrics tend to ignore load situations and thus result in overloaded clusters within the mesh network. Our work presents a novel approach for load-based routing. We developed load observation methods and proved their practical applicability within fixed IEEE 802.11a structures. The Extended Medium Observation (EMO)-metric is based on a sophisticated concept of an advanced channel observation process, in which topological and load sensitive characteristics become elicited, quantified and composed. A direct comparison of EMO-AODV, a modified version of AODV using EMO with conventional metrics show that EMO-AODV achieves a kind of traffic balancing that tends to smooth load peaks.