Research On Target Localization And Tracking Technology For WSN Based On CKF

Target localization and tracking is one of the key technologies on wireless sensor networks. The existing target localization and tracking algorithms have some disadvantages such as low accuracy, high energy cost and computational complexity. Therefore, how to improve the localization and tracking accuracy and reduce the computational complexity currently are hot research topics in the domestic and oversea countries. Node localization technology on wireless sensor networks is the basis of the target tracking technology, only after the whole nodes locate their own positions, can we use the wireless sensor networks to track the target efficiently. To do this, this thesis is mainly focus on the node localization and target tracking algorithms on wireless sensor networks based on cubature kalman filter.First of all, this thesis analyzes and summaries the related research status of target localization and tracking technologies on stationary and mobile wireless sensor networks at home and abroad, analyzes the performance evaluation standard of the localization algorithms and target tracking algorithms, lays a solid foundation for later research.Secondly, a detailed comparative analysis of four nonlinear filter algorithms are carried out, including extended kalman filter, unscented kalman filter, particle filter and cubature kalman filter. The accuracy of four nonlinear filter algorithms in one-dimensional nonlinear estimations are simulated and compared. The simulation results of UKF and CKF show that the accuracy of UKF is better than CKF under three dimension, but abc、3 dimension, CKF is better than UKF.Thirdly, in view of the wireless sensor network node location technology, to improve the positioning accuracy and reduce the computational complexity, this thesis proposes a distributed localization algorithm based on CKF on wireless sensor networks. Considering the node mobility, a distributed localization algorithm based on CKF on mobile wireless sensor networks is proposed. The simulation results show that the precision of the CKF algorithm is the same to the UKF, higher than the maximum likelihood estimation method, and the computational complexity of the CKF is lower than the UKF algorithm; the positioning accuracy of the localization algorithm based on CKF in MWSN is higher than localization algorithm based on CKF in WSN.Then, three kinds of target tracking models, including constant velocity, constant acceleration, and constant turning are introduced in this thesis. Besides, this thesis studies an interactive multiple model algorithm based on kalman filter and proposes a target tracking algorithm based on CKF on wireless sensor networks. The simulation experiments verify the effectiveness of this algorithm.Finally, the research work of this thesis is concluded and the future study topics are presented.