LOW-ANGLE RADAR TRACKING IN A NAVAL ENVIRONMENT USING A FORWARD-BACKWARD NONLINEAR PREDICTION METHOD

A low-angle tracking technique based on a novel high-resolution algori thm is presented. The new high-resolution algorithm, which is called t he forward-backward nonlinear prediction (FBNLP) method, replaces the linear predictor in the conventional forward-backward linear predictio n (FBLP) method with a nonlinear one to try to improve the tracking in a realistic naval environment. Real radar tracking data over the sea, which were recorded with a 32-element sampled aperture antenna on an ''over-water'' path, were used in this study. They were recorded in a multipath environment, i.e., one where, in addition to the customary d irect signal, an indirect signal was also received by the antenna syst em, and separation of these two signals was less than one standard bea mwidth. The performance of the FBNLP method is shown to be better than that of the modified FBLP algorithm for the data sets used in this pa per.