一种动态传感器网络自适应定位算法

An Adaptive Localization Algorithm in Dynamic Sensor Network

针对动态无线传感器网络节点移动性和已有算法定位精度较低的问题,在蒙特卡洛算法的基础上,提出了一种基于蒙特卡洛算法自适应定位算法.该算法结合了蒙特卡洛算法和增强型蒙特卡洛定位算法,解决已有定位方法很难快速收敛的问题;在滤波期间加入滤波加权以及Voronoi图和权值,通过其双重筛选机制来提高这一阶段的准确性;通过样本遗传交叉生成策略加速算法的重要性采样,从而促进算法的收敛,减少计算量,降低能耗.仿真实验显示,改进算法充分利用节点移动性以及邻居节点协作定位,能够更快速有效地完成定位,大大提高了定位精度.

In the light of the node mobility and low positioning accuracy of existing algorithms indynamic wireless sensor network, an adaptive localization algorithm is proposed based on MonteCarlo Algorithm （MCL）. The algorithm combines MCL and the enhanced Monte Carlo localizationto solve the problem of slow convergence speed of existing algorithms. During the filtering ofthe traditional algorithm, the filter weight and Voronoi map and weight value are added to improvethe accuracy of this stage by the dual screening mechanism. The algorithm is improved bythe sample genetic cross generation strategy to accelerate the importance sampling, thus to promotethe convergence, reduce the amount of computation and reduce energy consumption. Thesimulation results show that the improved algorithm makes full use of the node mobility as well asthe collaborative localization of neighbor nodes, can complete positioning quickly and efficientlyand improve the positioning accuracy greatly.