Robust Deployment for Data Collecting under Random Node Failures in Wireless Sensor Networks

The robust data collection is extremely important for the practical application of wireless sensor networks. Based on the underlying graph of wireless sensor networks, a system model to evaluate the robustness of the networks quantitatively to collect sensing data is constructed. The optimal solution is obtained with hill-climbing algorithm and Monte Carlo method. The result shows that the common sensor nodes should be clustered near the sink node when the total number of nodes becomes less or the failure coefficient is big, whereas they should be dispersed under the opposite conditions.