Sensor fusion for target detection and tracking

Abstract

Multipath can be a significant contributor to poor detection performance in radar ground moving target tracking systems. Recently, ground multisensor networks have been used to detect and track targets. Unfortunately, such networks can only be deployed in a limited area compared with the coverage possible using a radar. We show how, in those areas where multipath contributes to detection errors, the ground sensor network performance can be improved. Using integrated radar and ground sensor networks significantly improves the performance of radar target detection in a multipath environment. Such improvement in performance has implications for using more varied type of sensors to support the radar with statistics that complement the radar's receiver operating characteristic.