Association of multisite radar data in the presence of large navigation and sensor alignment errors

This paper discusses a technique for associating radar track data from two or more moving platforms in the presence of very large navigation errors and sensor misalignments. In practice, these errors and misalignments can be so large as to render conventional association/correlation algorithms ineffective. A novel approach to association is to view each platform's track file as a picture and apply two dimensional signal processing techniques to identify common features in the different pictures. This approach produces track association decisions which are unaffected by the size or direction of the navigation errors and by the size or sign of azimuth misalignments. This technique was tested using two 3D pencil beam surveillance radars separated by about 66 km. Navigation/alignment errors were artificially inserted into the track data bases and shown not to affect track association decisions, thus allowing an integrated surveillance picture to be produced even when navigation/alignment errors exceed typical target separations.