最小二乘拟合的蒙特卡罗移动定位算法研究
2018-08-06谭志梁丽文夏磊
谭志 梁丽文 夏磊



摘 要: 针对传统蒙特卡罗定位算法由于节点采样效率低导致的定位精度低、定位不准确的缺陷,提出一种改进的最小二乘拟合蒙特卡罗(LSFMCL)定位算法。该算法利用MBC算法优化采样空间,并利用最小二乘拟合节点运动轨迹,对节点位置进行预测,进一步得到最优采样区域,最后提出权值概念并利用预测节点的权值信息计算未知节点的位置。仿真结果表明,与传統算法相比,优化后的算法提升了节点的采样率,提高了定位精度,对于移动节点的定位具有更加广泛的应用前景。
关键词: 蒙特卡罗定位算法; MBC算法; 最小二乘法拟合; 移动节点定位; 运动轨迹; 权值概念
中图分类号: TN911.1?34; TP212.9 文献标识码: A 文章编号: 1004?373X(2018)15?0010?06
Research on least squares fitting Monte Carlo localization algorithm for mobile nodes
TAN Zhi, LIANG Liwen, XIA Lei
(College of Electrical and Information Engineering, Beijing University of Civil Engineering and Architecture, Beijing 100044, China)
Abstract: The traditional Monte Carlo localization algorithm has the defects of low positioning precision and inaccurate positioning due to the low efficiency of node sampling. Therefore, an improved least squares fitting Monte Carlo localization (LSFMCL) algorithm is proposed. The algorithm is adopted to optimize the sampling space by taking advantage of MBC algorithm, fit the node motion trajectory with least square method to predict the location of nodes, and then obtain the optimal sampling area. The weight concept is proposed to calculate the location of the unknown node by using the weight information of the predicted node. The simulation results show that, in comparison with traditional algorithm, the optimized algorithm can improve the sampling rate of the node and positioning accuracy, and has wider application prospect for the localization of the mobile nodes.
Keywords: Monte Carlo location algorithm; MBC algorithm; least square fitting; mobile node localization; motion trajectory; weight concept
0 引 言
无线传感器网络技术(Wireless Sensor Networks,WSN)作为21世纪发展的热点,具有很多别的技术所没有的优点,如监测范围广、运行成本低、网络自组织性好等。目前,传感器网络广泛地应用在智能家居、建筑防火、环境监测、军事侦察等领域[1]。
WSN技术的节点定位主要是指已知自身位置的节点通过某种测量方法来定位未知节点位置的方法。因此作为WSN热门技术之一的节点定位算法有不同的分类标准。根据是否需要测距分为基于测距的定位算法和无需测距的定位算法。典型的基于测距的定位算法包括TOA(Time of Advent)定位算法[2]、TDOA(Time Difference Of Arrival)定位算法[3]、 RSSI(Received Signal Strength Indication)定位算法[4]、AOA(Angle of Arrival)定位算法[5]等,而典型的无需测距的定位算法包括质心算法[6]、DV?Hop 算法[7]、Amorphous 算法[8?9]、APIT 算法[10]等。根据WSN中传感器节点是否能够移动,节点的定位算法分为静态定位(Static Localization)和动态定位(Dynamic Localization)两种类别。……
