一种基于粒子滤波的WSN自适应定位算法

An Adaptive Localization Algorithm Based on Particle Filter for Wireless Sensor Networks

针对WSN节点定位问题中基于动态采样的AMCB算法在某些定位时刻粒子数过低,粒子的最优继承得不到保证的问题,提出了改进算法AMCB*.为解决AMCB算法在锚节点密度较低,节点所处网络状况不佳时定位精度不高的问题,提出一种基于粒子滤波的WSN自适应定位算法——NAMCB,通过对节点所处的网络状况进行评估来设置合理的粒子数、定位方式与滤波条件.对80个节点连续定位125次的结果表明,NAMCB算法的平均定位误差为节点通信半径的0.344 2倍,平均采样次数为84次,相对于AMCB算法分别降低了12.26%与4.55%.

As the number of particles of the dynamic sampling based adaptive Monte Carlo boxed （AMCB）algorithm is too small at some localization time slots and the optimal inheritance of the particles can not be ensured,an improved particle filter localization algorithm named AMCB＊ was proposed. To solve the problem that the localization accuracy of the AMCB algorithm is low when the anchor node density is low and nodes are in bad network condition,an adaptive localization algorithm based on particle filter in WSN named NAMCB was proposed. NAMCB sets reasonable number of particles,localization mode and filtering condition based on the evaluation to the network condition of nodes at each localization time slot. In tests,when 80 nodes were continuously positioned 125 times, the average localization error of NAMCB was 0. 344 2 times of node communication radius,and the average sampling number was 84. Compared with AMCB,both were reduced by 12. 26% and 4. 55%respectively.