In this paper, we propose a new target tracking approach for wireless sensor networks (WSNs) by using the extended H-infinity filter. First, the extended H-infinity filter for nonlinear discrete-time systems is deduce...In this paper, we propose a new target tracking approach for wireless sensor networks (WSNs) by using the extended H-infinity filter. First, the extended H-infinity filter for nonlinear discrete-time systems is deduced through the Krein space analysis scheme. Then, the proposed extended H-infinity filtering algorithm is applied to target tracking in wireless sensor networks. Finally, experiments are conducted through a small wireless sensor network test-bed. Both experimental and simulation results illustrate that the extended H-infinity filtering algorithm is more accurate to track a moving target in wireless sensor networks than using the extended Kalman filter in the case of having no knowledge of the statistics of the environment and the target to be tracked.展开更多
基金supported by the National Science Foundation for Distinguished Young Scholars of China (No. 60825304)the Taishan Scholar Program of Shandong Province+1 种基金the National Basic Research Development Program of China (973Program) (No. 2009cb320600)the Open Project of State Key Laboratory of Industrial Control Technology in Zhejiang University (No. ICT1006)
文摘In this paper, we propose a new target tracking approach for wireless sensor networks (WSNs) by using the extended H-infinity filter. First, the extended H-infinity filter for nonlinear discrete-time systems is deduced through the Krein space analysis scheme. Then, the proposed extended H-infinity filtering algorithm is applied to target tracking in wireless sensor networks. Finally, experiments are conducted through a small wireless sensor network test-bed. Both experimental and simulation results illustrate that the extended H-infinity filtering algorithm is more accurate to track a moving target in wireless sensor networks than using the extended Kalman filter in the case of having no knowledge of the statistics of the environment and the target to be tracked.
