基于多尺度的EKF滤波融合定位跟踪算法  

The algorithm of location tracking by EKF filtering based on multi-scale

作　　者：田增山[1] 邢培基[1] 周非[1] 

机构地区：[1]重庆邮电大学移动通信技术重点实验室,重庆400065

出　　处：《电子技术应用》2008年第10期120-123,共4页Application of Electronic Technique

基　　金：教育部科学技术研究重点项目(NO.207097)

摘　　要：针对无线定位中快速移动目标的定位精度差的问题,提出了一种简单有效的定位跟踪算法。该算法将信号的多尺度分析方法与基于TDOA的蜂窝网定位技术相结合,基于某尺度上获得的单一的观测量,建立了一个新的多尺度的观测模型。基于新的多尺度观测模型,利用扩展Kalman滤波对目标进行定位跟踪,获得比在原始尺度上直接进行滤波定位跟踪更好的效果。通过仿真验证了该方法对提高蜂窝网无线定位精度的有效性。In view of the poor accuracy of cellular wireless location, one kind of simple and effective localization tracking algo- rithm is proposed. By combining the TDOA-based localization technology with the multi-scale analysis method of signal, a new multi-scale measured model based on a single measurement from a single scale is proposed. Based on the new multi-scale mea- sured model and Expanded Kalman Filtering, the estimate obtained by using the new algorithm is more accurate. A set of simula- tions is performed, and the results show the efficiency of the improved accuracy of cellular wireless location.

关 键 词：多尺度估计 到达时间差 定位跟踪 扩展Kalman滤波 观测模型 

分 类 号：TN953[电子电信—信号与信息处理] V448.2[电子电信—信息与通信工程]