Comparative Study of Bearing Only Tracking (BOT) System using Efficient Observer Trajectories

The main advantage of using passive target tracking is to identify the target state without exposing the presence of observing platform. The Bearing Only Tracking (BOT) is the field of underwater signal processing which is mainly focused on estimating the unknown position of target source by using the direction or angle only information. Therefore, observer can acquire the azimuth or angle information only; and cannot obtain the distance and velocity information about the target. For uniform motion targets, the target remains unobservable in the presence of stationary observing station and therefore the observing station must maneuver to track the unknown position and velocity of the target. The observability conditions for single moving observer are studied which is moving with constant velocity and acceleration using bearing only measurements. Keeping in view these observability conditions, the mathematical model for observer motion trajectory has been developed in this paper with finite number of constant velocity segments which are termed as Legs. These legs are straight line segments with observer course change in the beginning of each time interval. Secondly, the mathematical model for the sinusoidal or coordinated turn observer motion has been defined which changes the angular velocity to observe the unknown target. Thus, these two observer trajectories primarily fulfil the observability criteria in case of maneuvering observing station. The underwater BOT system has non-linear measurement model and therefore, we cannot use the standard estimation and filtration methods to estimate the target states. Thus, to resolve this non-linearity, the EKF algorithms are used throughout the paper. The simulation experiments are performed for above observer trajectories and it is analyzed that such observer trajectories have got higher observability in the form of increased bearing rate and ultimately in reducing estimation errors.