卡尔曼滤波在无线传感器网络节点定位中的应用

Application of Kalman filter algorithm to node localization in wireless sensor networks

为了减小无线传感器网络节点定位中节点测距误差和定位算法自身引入误差的积累对定位精度的影响,建立了适用于无线传感器网络的卡尔曼滤波模型。采用渐消自适应卡尔曼滤波对基于距离的初始定位算法进行求精,通过一跳节点间的相互制约,在全局范围对未知节点进行定位;基于卡尔曼滤波模型,对无线传感器网络的分布式算法和集中式算法的计算量进行了研究,得出计算量与节点数目的关系。研究结果表明,卡尔曼滤波能够有效提高节点的定位精度,尤其适用于网络节点密度小、信标节点比例低的情况。

A Kalman filter model was established for wireless senor networks to reduce the influence of the accumulation of the location error and the error introduced by localization algorithms on the location precision in node localization. The algorithm for initial estimates based on range was refined by a fading Kalman filter. By the restraint between neighboring nodes, the locations of unknown nodes were computed. The computation of the centralized algorithm and the distributed algorithm was researched based on the Kalman filtering model, and the relationship between the computation and the number of the nodes was obtained. The simulation results showed that the localization precision was effectively improved by the Kalman filter. The Kalman filter is especially applicable to the network in which both the node density and the beacon ratio are low.