An exploration of geographic routing with k-hop based searching in wireless sensor networks

We explore the asymptotic performance of existing geographic routing with a utilization of k-hop neighborhood information. The reachability from source to sink improves as we integrate more information into the routing decision. It is observed that the average number of hops required decreases significantly from the 1-hop to 2-hop based searching. This indicates an attractive tradeoff between the performance enhancement and system complexity. As simple greedy forwarding is insufficient in lossy wireless environment, we propose a new metric incorporating advance in distance and link quality to handle unreliable communication links. Simulation result has verified its effectiveness and superiority over the conventional simple greedy method. Besides, a generalization to k-hop based routing and the resulting performance are presented. Results show that with the multi-hop based searching, in both the deterministic and probabilistic lossy radio models, there is a good improvement in the number of transmissions required from end to end, which can indicate potential improvement in the routing delay and energy efficiency in transmissions.