一种强背景噪声下的WSN目标定位算法

为了进一步提高无线传感器网络(WSN)目标定位解算精度,提出了一种改进的Cubature粒子滤波(ICPF)定位算法.该算法运用最小二乘法估计移动目标当前初始时刻的位置,使用Cubature卡尔曼滤波和Gauss-Newton迭代法来充分利用测量更新后的状态最新信息,精确设计目标状态重要性密度函数,为粒子滤波提供相应的建议分布,从而能够更加有效改善粒子滤波器的性能.仿真实验结果证明,提出的改进算法在强背景噪声下能有效提高定位精度且收敛性增强,其性能优于标准粒子滤波(PF)、扩展粒子滤波(EPF)及Unscented粒子滤波定位算法(UPF). 

An Improved Localization Algorithm in Wireless Sensor Networks Under Complex Background Noise

To further improve the precision of target localization in the wireless sensor network(WSN), a novel Cubature particle filter algorithm was proposed. Firstly it utilized least square method to estimate the initial target position, the Cubature Kalman filter and Gauss-Newton rule were used to generate more accurate importance density function subsequently by virtue of the measurement updated state variable, which provided proposal distribution for particle filter. The simulation experiment results demonstrate that the proposed algorithm can achieve better locating accuracy and convergence speed under complex background noise, compared with standard PF、 EPF and UPF.