Improved handset tracking using Kalman filter algorithm aided by angle-spread information from a smart antenna array

The paper addresses the problem of tracking a cellphone handset in a multipath environment using two or more base stations equipped with smart antenna arrays in conjunction with an extended Kalman filter. The Kalman filter is very sensitive to both the measurement variances and trajectories of the target handset. However, by suitably tuning the Kalman filter, the error in the position estimate of the handset can be significantly reduced. A new method is proposed to provide improved tracking of cellphone handsets in a typical urban multipath environment. In this new method, an estimate of the measurement errors in the angle of arrival measurement due to multipath is obtained from an antenna array at each base-station using a modified MUSIC algorithm and this is used to tune the extended Kalman filter. The proposed technique can be used to improve the tracking performance for non-linear trajectories.