The Research On The Data Fusion Algorithm In Wireless Sensor Network

With the development of low-power wireless communication technology as wellas sensors miniaturization, wireless sensor network has become the hot research field.WSN exists data redundancy phenomenon in the data transfer process, and the sensornodes energy, processor power, storage capacity and bandwidth are subject torestrictions, better use of data fusion algorithm to reduce the amount of network datatransmission, extended network lifetime imminent. In this paper, how to design toestimated high precision, good real-time distributed fusion algorithm as the mainresearch content.Kalman Filter (KF) is one of the generally applicable distributed fusion algorithm.Consistency of policy objectives through the exchange of information betweenneighboring nodes and distributed iterative weight, making the status of all nodes inthe network to converge. In this paper, the classical Kalman filter algorithm andconsistency algorithm combined to obtain a detailed analysis of the processes,characteristics and simulation of the algorithm. The simulation results show that, incertain conditions of the network scale and the sensor communication radius, thealgorithm can be approximated to achieve the estimation accuracy of the centralizedKalman filter algorithm with fusion center.Based on the idea of classic Kalman filter algorithm combined with consistency foronline system, tracking for nonlinear systems monitoring,the consistency algorithmand extended Kalman filter algorithm combined, this paple analyse the characteristicsand performance of this algorithm. The simulation verify the accuracy andperformance of the algorithm.