非视距环境下WSN移动节点定位算法研究

发布时间：2018-05-13 18:13

本文选题：无线传感器网络 + 非视距　；参考：《东北大学》2014年硕士论文

【摘要】：无线传感器网络(Wireless Sensor Network, WSN)是一门新兴的信息采集与处理技术,具有极为广阔的应用前景。移动节点定位作为WSN的关键技术,是无线传感器网络研究的热点问题之一。目前的研究通常仅考虑视距状态(Line-of-sight, LOS),但在实际应用中,信号的非视距(Non-line-of-sight, NLOS)传播现象是普遍存在的,并将导致定位算法的精度大大降低。本文对NLOS环境下的节点定位问题展开了深入的讨论和研究,目的在于提高NLOS环境下移动节点的定位精度。本文分析了NLOS距离测量值的残差特性,并提出严格残差选择机制以对距离测量值进行状态鉴别。由于距离测量值通常既包含LOS测量值又包含NLOS测量值,因此,充分利用LOS测量值进行定位可以有效提高算法的定位精度。本文首先应用扩展卡尔曼滤波(Extended Kalman Filter, EKF)算法的线性回归模型生成测量值的残差,并利用LOS测量值与NLOS测量值的残差差异完成测量值状态的准确鉴别。仿真结果表明,在NLOS环境下,采用严格残差选择机制进行测量值的状态鉴别,结合变节点EKF算法可以得到较高的定位精度。基于M-估计算法定位思想,本文提出了一种移动节点鲁棒定位算法。通过分析NLOS残差的统计特性,提出了一种基于最邻近估计的变量核密度估计算法,以估算残差的概率密度函数。然后结合M-估计的定位思想,提出了一种基于变量核密度估计的移动节点定位算法。实验结果表明,该算法克服了M-估计算法需要模型匹配与人工调整参数的局限性,实现了对不同环境下NLOS误差的抑制。考虑到NLOS误差的特性,本文提出了一种基于投票选择机制的概率数据关联算法。利用NLOS误差标准差大于测量误差标准差的特性,然后结合高频测距数据处理思想,提出了一种基于投票选择机制的数据处理算法对距离测量值进行筛选,并保留可靠的测量值。在此基础上,提出了一种改进的概率数据关联算法对经投票筛选后的多个测量值进行数据融合,最后应用基于参考节点选择的线性最小二乘估计算法计算出移动节点的位置。该算法有效地削弱了各种类型的NLOS误差,提高了移动节点的定位精度。本文系统地研究了非视距环境下WSN的移动节点定位算法,并通过一系列仿真实验与现场实验对所提算法进行分析。实验结果证明了本文所提算法在NLOS环境下均具有较强的鲁棒性和较高的定位精度。
[Abstract]:Wireless Sensor Network, WSN) is a new technology of information acquisition and processing, which has a wide application prospect. As a key technology of WSN, mobile node location is one of the hot issues in wireless sensor networks. At present, only Line-of-sight (LOSN) is considered in the current research. However, in practical applications, the phenomenon of non-line-of-sight (NLOSs) propagation of signals is widespread, which will result in the reduction of the accuracy of the localization algorithm. In this paper, the problem of node location in NLOS environment is discussed and studied in order to improve the positioning accuracy of mobile nodes in NLOS environment. In this paper, the residual characteristics of NLOS distance measurements are analyzed, and a strict residual selection mechanism is proposed to identify the state of the distance measurements. Because the distance measurement value usually includes both the LOS and NLOS measurements, the localization accuracy of the algorithm can be improved by fully utilizing the LOS measurement values. In this paper, the linear regression model of extended Kalman Filter, EKF) algorithm is first used to generate the residual error of the measured value, and the state of the measured value is accurately identified by the difference between the measured value of LOS and that of the value of NLOS. The simulation results show that in the NLOS environment, the strict residual selection mechanism is used to identify the state of the measured values, and the variable node EKF algorithm is used to obtain higher positioning accuracy. Based on the idea of M- estimation algorithm, a robust location algorithm for mobile nodes is proposed in this paper. By analyzing the statistical properties of NLOS residuals, a variable kernel density estimation algorithm based on nearest neighbor estimation is proposed to estimate the probability density function of the residuals. Then, a mobile node location algorithm based on variable kernel density estimation is proposed based on the idea of M- estimation. The experimental results show that the algorithm overcomes the limitation of the M- estimation algorithm which requires model matching and manual adjustment of parameters, and realizes the suppression of NLOS errors in different environments. Considering the characteristic of NLOS error, a probabilistic data association algorithm based on voting selection mechanism is proposed in this paper. Taking advantage of the fact that the standard deviation of NLOS error is larger than the standard deviation of measurement error, and combining with the idea of high frequency ranging data processing, a data processing algorithm based on voting selection mechanism is proposed to filter the distance measurement value and keep the reliable measurement value. On this basis, an improved probabilistic data association algorithm is proposed for data fusion of several measured values selected by voting. Finally, the location of mobile nodes is calculated by using the linear least square estimation algorithm based on the selection of reference nodes. The algorithm effectively weakens all kinds of NLOS errors and improves the location accuracy of mobile nodes. In this paper, the mobile node location algorithm of WSN in the non-line-of-sight environment is systematically studied, and the proposed algorithm is analyzed through a series of simulation experiments and field experiments. The experimental results show that the proposed algorithm is robust and accurate in NLOS environment.
【学位授予单位】：东北大学
【学位级别】：硕士
【学位授予年份】：2014
【分类号】：TN929.5;TP212.9

【相似文献】