摘要
由于MDSMAP算法中测量距离与节点坐标之间的非线性的关系,从而无法快速得到节点位置信息,而且由于MDSMAP算法在定位精度和定位误差等方面存在着不足,因此在此基础上引入扩展卡尔曼滤波对MDSMAP进行求精.以MDSMAP算法得到的定位位置为初始数据,利用SVD分解,得到定位坐标,然后同多建立EKF模型,对节点位置进行迭代,从而可以得到精度更高的位置信息.最后以Matlab软件做为仿真平台,在网络覆盖范围为100*100、通信距离为25、30和50的条件下,对改进前后的算法进行仿真,结果显示10次的仿真在精度和误差上都提高了5%.
Because of the non-linear relationship between the measurement of distance and node coordinates in the MDS-MAP algorithm,it is impossible to get the location information of nodes quickly.Moreover,owing to the disadvantages of the MDS-MAP algorithm in positioning accuracy and positioning errors,this paper introduces Extended Kalman Filter to refine the MDS-MAP algorithm.The location obtained by MDS-MAP algorithm as the initial data,the accurate location was got by using SVD decomposition and an EKF model was established.The effect of the localization error and location coverage of anchor nodes are improved by 5%when the communication distance is 25,30 and 50,and the communication coverage is 100*100.
作者
范海红
沈水金
FAN Haihong;SHEN Shuijin(Department of Management and Information,Zhejiang Post and Telecommunications College,Shaoxing,Zhejiang 312000;Department of Mathematics,Physics and Information Science,Shaoxing University,Shaoxing,Zhejiang 312000)
出处
《绍兴文理学院学报》
2019年第8期61-67,共7页
Journal of Shaoxing University
基金
浙江省教育厅项目“无线传感器节点自定位算法的研究”(Y201738882)
浙江省高等学校访问工程师校企合作项目“无线传感器网络节点定位算法的研究和应用”(FG2017218)
关键词
无线传感器网络
节点定位
卡尔曼滤波
仿真
wireless sensor network
localization
Extended Kalman Filter
simulation
