EasiDSlT: A Two-Layer Data Association Method for Multitarget Tracking in Wireless Sensor Networks

The technology of multitarget tracking (MTT) has been widely and deeply researched in many fields, such as the radar system and wireless sensor networks (WSNs). However, how to develop a lightweight data association algorithm in a decentralized way is still a challenge, particularly considering the fact that WSNs are resource constrained. This paper presents a two-layer data association method for MTT applications, which are based on low-cost WSNs. To improve the association accuracy of the first layer of the data association, this paper proposes a lightweight reasoning method based on the evidence theory. The example analysis indicates that it can also handle the problem of highly conflicting information fusion. The second layer adopts a Bayesian filtering algorithm. By adoption of the two-layer data association, the computation cost of data association in the MTT technology is balanced in intracluster nodes. Simulation experiments show that the data association algorithm has great performance.