Dependable data aggregation on cluster-based wireless sensor networks

In the paper we propose a dependable and efficient data aggregation scheme based on fault map that is constructed by estimated fault probability using Bayesian Belief Network (BBN). Cluster head will forward aggregated event depend on the total dependability of collected event in which we assign each detecting node a dependence weight. The dependence weight of sensor node is mapped from its fault probability estimated by cluster head. We propose a mapping function to map node's fault probability into a dependence weight. Cluster head accumulates the dependence weight instead of number of source node and transmits the aggregated event when the threshold has reached. The simulation result shows that the approach even though take some extra delay time, it will increase the credibility and dependability of the aggregated event in the naturally unreliable wireless sensor network.