An Improved Probabilistic Data Association Algorithm in Wireless Sensor Networks

Abstract

Data association algorithm is considered as an important part in modern target tracking systems. Although traditional probabilistic data association algorithms are used in various areas and perform well, there are numbers of disadvantages. In this article we introduce an advanced algorithm that perform differently depends on measurements of tracking gates, filtering out weak correlated or uncorrelated data to reduce the expanse while improve the accuracy. In the last part, the algorithm is compared with probabilistic data association algorithms. The comparison shows that the advanced algorithm reduces the number of relative position error as well as the computation time at a certain point of level, which proves the effectiveness of the algorithm.