Simultaneous Localization And Tracking In Wirless Sensor Networks

The Wireless sensor networks(WSNs) is a novel technology to obtaining and processing information. Meanwhile WSNs is considered as one of the most important technology in 21 st century. WSNs can be widely used in many fields including: national army defense, environment monitoring, medical and health care, object tracking, space exploring and so on. Target tracking is one of the most important applications of WSNs. It is usually assumed that the knowledge of the sensor nodes’ position is known precisely. However, practically nodes are randomly deployed without prior knowledge about their own position. To the problem of simultaneous localization and tracking(SLAT), this paper comprehensively discussed and studied the centralization data fusion and distribution data fusion in simultaneous localization and tracking problems respectively. At the same time, a new distributed fusion method is proposed in this paper. We make a compare of the proposed algorithm and other method and conduct the relevant simulations.At first, this paper make a introduction of the classical filtering algorithm in WSNs, such as Kalman filter(KF), extended Kalman filtering(EKF), unscented Kalman filtering(UKF), and interactive mutiple model(IMM). Combined with the recent research hotspot on wireless location tracking technology, this thesis focuses on the centralization date fusion of Interactive Multiple Model based on Kalman Filter(IMM-UKF), Interactive Multiple Model based on Extended Kalman Filter(IMM-EKF), Interactive Multiple Model based on Unscented Kalman filter(IMM-UKF). The simulation analyzed each algorithm’s advantages and disadvantages in the maneuvering situation and non-maneuvering situation. It shows that IMM based on UKFs has better accuracy in both localization and tracking, as well as higher robustness when it is compared with other methods.Aiming at the problem of simultaneous localization and tracking in wireless sensor networks, a distributed fusion algorithm is proposed in this paper, in which the interactive multiple model based on unscented Kalman filter is adopted for local estimator. And to fuse the local estimates of target track, the internal ellipsoidal approximation fusion(IEAF) is used at the fusion center. Monte Carlo simulation results show that the proposed approach has higher fusion accuracy of 33.56% than the covariance intersection(CI), and obtaining fairly accurate estimates of the node location.At the end of the article, we not only sum up the main points of SLAT problem and make a summarization, but also we make a outlook of the future research orientation of the problem of SLAT.