基于Monte Carlo的无线传感器网络定位算法

无线传感器网络节点自身定位技术是无线传感器网络关键技术之一。针对目前各种定位算法存在定位精度较低的问题,提出了一种基于Monte Carlo方法的定位算法,该算法利用粒子到锚节点的距离计算各粒子的权值,通过滤波不断更新粒子的集合,使粒子收敛到未知节点的位置。对非视线情况、不同锚节点个数、迭代次数及粒子数进行了定位过程仿真,并和极大似然估计定位算法进行了定位结果比较。结果表明:该算法充分利用了对节点位置估计的有效信息,一定程度上抑制了非视线误差的影响,定位精度高,稳定性好。

Localization algorithms of wireless sensor networks based on Monte Carlo method

Wireless sensor networks node self-location plays a critical role in the application of wireless sensor networks.But high inaccuracy is still a problem in its application.To solve this problem,a new localization algorithm based on Monte Carlo is proposed.By computing the power value with particle-to-anchor node distance,algorithm converges to the unknown location through the constant particles group updating.It has simulated at the occasion of no line of sight,different quantity anchor nodes,iteration and particles.And it also compared with maximum likelihood estimation localization algorithm.The result shows that the localization algorithm makes full use of the node's localization information;and reduces affection of no line of sight with high accuracy and of well robustness.