Multi-sensor, Multi-target Localization And Tracking Based On Passive Radar

Passive radar can break through the limitations of active radar in the stealth technology and low altitude penetration, so the radar system has been widely developed in recent years. Passive radar system has been developed from the single sensor system to the multi-sensor system, and in the case of multi-target tracking, multi-sensor system can enhance the adaptability of multi-target tracking by using data fusion technology. In this paper, the problems of multi-target localization and tracking have been mainly studied, and the key technologies of multi-target localization and tracking have been discussed in the multi-sensor passive radar system.In this paper, a data processing program has been designed in the passive radar system, which consists of three transmitting stations and one receiving station. The program includes data preprocessing, point fusion and track processing. Data preprocessing mainly deals with the measured data by track processing in the range-doppler dimension, thereby eliminating false alarm points, smoothing the measured data and reducing the amount of data processing calculation. The main purpose of points fusion is to translate the measured data in the spherical coordinate to the data on the XY plane, thus these points on the XY plane are classified which are corresponded to the different targets. The module includes exhaustive pairing, TOA localization, clustering and fusion. First, the exhaustive pairing is to pair the exhaustive measured data from the differernt transmitting stations. Second, we use the TOA localization to translate the measured data in the spherical coordinate to the data on the XY plane. Third, the clustering and fusion which used in this paper can classify these points on the XY plane and fuse the points in one class. The points fusion module analyzes TOA localization accuracy factors and selects the coarse clustering used for the points classification. The main purpose of the target track processing is to deal with the points which have been fused, the process includes track starting, point-track association, track forecast and track demise. The Kalman filtering algorithm is selected to be the method of the track process.Finally, the multi-sensor multi-target localization and tracking have been simulated and analyzed by the designed data processing program, and also measured data has been used to verify the performance of the tracking process.