一种基于修正卡尔曼滤波的蜂窝定位算法

摘要

在蜂窝网络中,非视距(Non-line-of-sight,NLOS)误差是影响定位精度的关键因素.本文以卡尔曼滤波和Greenstein模型为基础,通过判决鉴别出NLOS误差,借助其均值修正卡尔曼预测过程中产生的异常新息,直接消除NLOS误差,再利用多项式平滑滤除测量误差,重构到达时间(Time-of-arrival,TOA)测量值;接着对卡尔曼滤波的量测矩阵进行修正,用重构的测量值对移动台精确定位.仿真结果表明,该算法能够有效地抑制NLOS误差,提高NLOS传播环境下的定位精确度,在一定程度上满足了E-911的定位需求.%In cellular network, non-line-of-sight(NLOS) error is the key factor that affecting positioning accuracy. Based on Kalman filter and Greenstein model, NLOS error is identified by judgment and directly mitigated with its mean modifying new abnormal information caused during Kalman prediction process. Then the measurement error is eliminate with polynomial smoothing, thus reconstructing the time-of-arrival (TOA) measurements. Whereafter, it modifies the measuring matrix and obtains the locating results using the reconstructed measurements. The simulation results show that the proposed method can mitigate NLOS error effectively and improve the location accuracy in the condition of NLOS propagation. It meets the demands of E-911 to some extent.