Characterizing the traffic load distribution in dense wireless sensor networks

Traffic load is typically not evenly distributed over the sensor nodes in a wireless sensor network (WSN). Understanding the traffic load distribution can guide the network-wide energy allocation, the design of routing algorithms, and the optimization of the node deployment in WSNs. This article considers a dense WSN with nodes uniformly distributed in a disk sensing area, where the expected traffic load distribution over the sensor nodes as a function of their distance from the sink has been derived. Further, the effects of the network scale and routing strategy on traffic load distribution are also investigated. The expected traffic loads beared by individual sensor nodes are found to be in direct proportion to the radius of the network and inversively proportional to the routing hop length but it is independent of the network density. In addition, a heuristic multipath routing algorithm is found to be capable of reducing the traffic load variance experienced by neighboring or symmetrically deployed sensor nodes. The results presented in this article are verified through extensive simulation experiments.