Skip to Main Content
A major challenge in designing MAC protocols for wireless sensor networks (WSN) is the uncertainty about the traffic offered by network, which usually forces conservative assumptions leading to a degradation in throughput and delay performance. Traffic estimation is discussed here in the context of the distributed detection WSNs (DD-WSNs). We approach this issue by first showing that the traffic has a Poisson distribution via stochastic geometry tools. Then the traffic is estimated via two algorithms, the least conditional maximum a priori (lcMAP) estimator and the regularized maximum likelihood estimator (rMLE). To measure the correlation between supplied communication resources and needed resources by the WSN, we propose the supply demand ratio (SDR) as a metric. Simulation results shows that both estimators achieve a performance close to the optimal MAP estimator under low channel SNR, hence transmission energy can be saved. Furthermore, the rMLE achieves the optimal SDR via choosing regularization factor value.