在蜂窩移動定位中，非視距波(NLOS)是影響定位精度的主要因素，NLOS的傳播會給距離測量值增加較大的正性誤差。傳統的線性濾波器主要是針對高斯分布的誤差，并不能有效地處理NLOS誤差。該文提出一種非線性LMS算法TOA測量結果進行重構來消除NLOS的影響。該算法將NLOS誤差視作沖激性噪聲，通過對LMS算法中的代價函數的非線性化來抑制TOA測量值中的NLOS誤差，仿真結果證明了該算法的有效性。
關 鍵 詞 非線性LMS; 非視距; 無線定位; 到達時間

Abstract The major problem to make precision mobile location is the nonline of light (NLOS) effect. It adds positive error to time of arrival (TOA) measures, the linear filter can not effectively deal with the NLOS error, because the linear filter mainly bases on the Gaussian noise. In this paper, a nonlinear LMS algorithm is brought forward to mitigate the NLOS effect on location precision. It regards NLOS error as the impulse noise and reconstructs the TOA measures by using the non-linear objective function. The simulation results justify the proposed algorithm.
Key words nonlinear LMS; nonline of sight; radio location; time of arrival