Research Of The Algorithm Based On TDOA For Wireless Sensor Network Location

As a key technology of wireless sensor network, the node localization is the foundation and the core technology of many applications, which plays a crucial role in wireless sensor network monitoring activity.Node localization algorithm based on TDOA is widely used at present, there are two main error sources in this kind of algorithm: TDOA ranging error caused by NLOS and the error of the nonlinear equations. This issue will research special vehicle’s localization in airport airfields. The ground segment of airfield includes runways, taxiways, aprons, etc. Because of the large number of aircraft, special vehicles, this area exist a serious NLOS error and the NLOS error model can’t be determined here. The algorithm of wireless sensor location based on TDOA is improved for Kalman filter algorithm to effectively inhibit TDOA ranging error caused by NLOS. By studying the error characteristics, estimates are subject to the introduction of new interest threshold for large NLOS error will cause serious impact on subsequent estimates, while for NLOS error can’t be avoided, correction factors are introduced so that Kalman filter algorithm is improved and we can get a more accurate estimate.In the algorithm of wireless sensor network based on TDOA, least-squares algorithm complexity low but it’s inferior in NLOS error. Taylor series expansion algorithm can effectively suppress NLOS error by iterations which still influenced by the initial estimate badly. Aiming at the shortcomings and problems of the two algorithms, this paper gives a combination of them. Meanwhile, as the driving characteristics of the airport vehicles constraints of the algorithm, it get the initial estimates of the least squares algorithm and set centroid of this estimate field as Taylor algorithm’s estimate to iterate. The results show that the improved algorithm has higher positioning accuracy.