无线传感器网络DV-Hop定位算法的优化研究

发布时间：2018-07-20 13:38

【摘要】：近年来,与无线传感器网络定位技术相关的应用越来越多,从而对无线传感器网络定位算法的研究也就显得很重要了,本文重点研究了无线传感器网络中基于非测距的DV-Hop定位算法,并提出了如下3种优化方案来提高节点定位精度:(1)针对节点分布不均匀引起的误差,本文提出基于区域划分锚节点移动的DV-Hop定位算法。即在定位过程中,先将目标区域平均划分成4个子区域,然后引入可移动的锚节点,通过3种不同的移动方法对锚节点进行移动再部署,使锚节点覆盖的范围尽可能大,最后再用传统DV-Hop算法进行定位。其中,3种不同移动方法分别为:锚节点均匀部署、锚节点按密度部署、锚节点按间隔阈值部署;(2)针对传统最小二乘法计算过程中带来的误差,本文提出基于自适应权重与学习因子同步变化的粒子群优化的DV-Hop定位算法。即用未知节点最短跳数的倒数对估计距离与实际距离的差值进行加权,并将平方后的值作为粒子群的适应度函数,同时引入自适应权重和同步变化的学习因子思想来克服粒子在寻优过程中容易陷入局部最优值、早熟收敛等问题,并用改进后的粒子群算法求解未知节点的位置;(3)将前两种算法进行结合,本文提出了基于APMAN-AdaCPSO优化的DV-Hop定位算法,利用AdaCPSO算法对APMAN-DV-Hop算法的定位结果进行优化。上面3种优化算法均在MATLAB平台下进行了仿真,仿真结果表明,改进算法的定位误差均比原始的DV-Hop算法要小,提高了节点的定位精度。
[Abstract]:In recent years, there are more and more applications related to wireless sensor network positioning technology, so the research of wireless sensor network location algorithm is also very important. This paper focuses on the wireless sensor network based on the DV-Hop location algorithm based on distance measurement, and proposes the next 3 optimization schemes to improve the location accuracy of nodes: (1) In view of the error caused by uneven distribution of nodes, this paper proposes a DV-Hop location algorithm based on the regional partition of anchor node movement. In the process of positioning, the target area is divided into 4 subregions, then the movable anchor nodes are introduced, and the anchor nodes are moved and redeployed through 3 different moving methods to cover the anchor nodes. The scope is as large as possible, and then the traditional DV-Hop algorithm is used to locate. Among them, the 3 different methods are as follows: the anchor nodes are deployed evenly, the anchor nodes are deployed according to the density, the anchor nodes are deployed at the interval threshold, and (2) this paper proposes that the adaptive weights are synchronized with the learning factors according to the error caused by the traditional least square method. The DV-Hop localization algorithm of particle swarm optimization is used to weigh the difference between the estimated distance and the actual distance by the reciprocal of the shortest hop number of the unknown node, and take the value of the square as the fitness function of the particle swarm, and introduce the learning factor of adaptive weight and synchronous change to overcome the easy fall of the particle in the process of optimization. Local optimal value, premature convergence and other problems, and using improved particle swarm optimization algorithm to solve the location of unknown nodes; (3) the first two algorithms are combined, this paper proposes a APMAN-AdaCPSO optimization based DV-Hop positioning algorithm, using the AdaCPSO algorithm to optimize the localization results of the APMAN-DV-Hop algorithm. The above 3 Optimization algorithms are all in MATLAB The simulation results show that the location error of the improved algorithm is smaller than that of the original DV-Hop algorithm, and the positioning accuracy of nodes is improved.
【学位授予单位】：南京邮电大学
【学位级别】：硕士
【学位授予年份】：2016
【分类号】：TN929.5;TP212.9

【相似文献】