Study Of Detection And Tracking For Radar

Detection and tracking are the basic functions which most of the radars must have, and this dissertation is mainly concerned about those two parts. The radar uses preprocessing techniques to reject undesired echoes before target detection. The widely used preprocessing techniques include matched filter, Moving Target Indication (MTI) filter, and coherent integration. After the preprocessing, the Constant-False-Alarm Rate (CFAR) technique is used to detect the moving targets at a constant false-alarm rate.Target tracking is the processing of measurements which in order to maintain an estimate of the target state, such as position, velocity and so on. The data association is done firstly for the measurements which falling within the validation region, then the state estimation, and finally updating the states of the targets. State estimation is completed by the Kalman filter, and data association is solved by the Probabilistic Data Association (PDA) approach when tracking a single target in clutter, or by the Joint Probabilistic Data Association (JPDA) approach when tracking multiple targets in clutter. The key to the JPDA algorithm is the evaluation of the probabilities of association, and this needs to obtain the feasible matrices from a validation matrix which becomes a difficult problem. We bring the concept of state codes to the creating process of feasible matrices, and record the needed historical information on a matrix, and consequently the programming implementation becomes greatly easier.