后验Monte Carlo-Gaussian采样粒子滤波WSN网络定位*

环境因素导致无线传感器网络定位存在噪声影响,实质上是非平滑的非线性问题,针对传统粒子滤波算法在处理该问题时精度不高的缺点,提出一种基于后验泊松分布的Monte Carlo-Gaussian重采样粒子滤波算法的无线传感器网络定位算法。首先,基于粒子滤波算法,借鉴扩展卡尔曼滤波算法采用近似后验高斯分布思想,设计了后验泊松分布Monte Carlo-Gaussian重采样粒子滤波器。其次,采用该滤波器设计实现了无线传感器网络定位算法,解决了非平滑非线性的噪声干扰定位问题。最后,分别对滤波器和定位算法的性能进行了对比仿真实验,结果验证了所提算法的有效性。

The posterior Monte Carlo Gaussian importance sampling par-ticle filter algorithm for Wireless Sensor Networks localization

Environmental factors cause localization in wireless sensor network exists noise effects, which is essentially nonlinear problem of non smooth, in view of the traditional particle filter algorithm in dealing with the issue of the disadvantage of low accuracy, the posterior Monte Carlo Gaussian importance sampling particle filter algorithm for Wireless Sensor Networks localization was proposed. Firstly, based on particle filter algorithm, the posterior Poisson distribution Monte Carlo-Gaussian re-sampling particle filter was designed with the thought of extended Calman filter algorithm and the approximate posterior distribution. Secondly, using the filter design to implement wireless sensor network localization algorithm, which solves the problem of non smooth nonlinear noise interfer-ence location. Finally, simulation experiment was carried respectively on the filter performance and the localization accuracy comparation, the results verify the effectiveness of the proposed algorithm.