A Fast Data Fusion Algorithm Based on Matrix Analysis for Target Recognition in Sensor Networks

Abstract

Data fusion always be used to solve the problem of target recognition in wireless sensor networks, and Dempster-Shafer inference have been proven one of effective methods. However, the high computation complexity of D-S evidence combination usually prevent it to be used in low ability sensor networks directly, especially it may violate the requirements for real-time processing and synchronization of sensor networks. To increase the speed of target recognition in wireless sensor networks, a fast data fusion algorithm based on matrix analysis is proposed in this paper, which inherits the idea of D-S evidence theory. The algorithm holds the same recognition capability as D-S evidence combination formula, but reduces time complexity. This conclusion has been confirmed by simulation.