Data fusion for underwater target tracking

Underwater manoeuvring target rarely tracked using bearings-only measurements available from Hull mounted array (HA) without a proper manoeuvre by the observer. This problem is solved by administering data fusion techniques on bearings available from towed array and HA. Song and Speyer´s and Galkowski and Islam´s modified gain bearings only extended Kalman filter is exploited for estimation of target motion parameters. Online pre-processing is carried out to reduce the amplitude of the noise, compute the estimated bearings if the bearing measurement is not available and to find out variance of the noisy measurement which is used in Kalman filter. The spurious measurements are made invalid. The performance evaluation of the algorithms is done in Monte-Carlo simulation and results obtained for two typical geometries are presented.