Splitting Algorithm for Cases with Single Detection from Sonar Data for Two Tracks

Abstract

The purpose of this paper is to analyze data association problem in sonar applications. Bearings only tracking algorithms are used because the range information is distorted in water. A method is proposed to associate detections with established sonar tracks using the distribution of the received signal. Using a single detection when two target's bearings are close to each other, may result in poor tracking capability. Splitting a single detection from two targets is a new method which has the potential to keep weaker target's track alive. This method is compared with PDA (probabilistic data association) filters. Monte Carlo simulation results are also given