Target tracking using matched-field processing

Abstract

The objective of this paper is to illustrate the use of matched-field processing (MFP) for tracking low signal-to-noise ratio targets moving linearly and at constant speed. The input to the tracker consists of the positions and power of the largest peaks on the MFP ambiguity surface. These largest peaks usually include the match at or near the source position even at low signal-to-noise ratio. An exhaustive search for the best matching track over all possible target tracks (that is allowing varying speed and heading) is beyond the scope of today’s computers for any realistic search region. In this paper, an efficient algorithm is described based on examining the average Bartlett statistic along a set of linear tracks that connect only the largest peaks. This set was restricted to the physically possible tracks to further reduce the number to be examined. Examples of the ambiguity surfaces and the probability of examining the true track are given. The algorithm performance is a function of the scenario, signal-to-noise ratio, number of ambiguity surfaces, and number of peaks examined on each surface. It is shown that if the true target track is one of those examined its Bartlett statistic is almost certainly maximum. This efficient tracking requires only modest computing beyond that required to generate the ambiguity surfaces.