Tracking Algorithm Based On The Sequence Detection Of Weak Target Detection Before

With the development of modern military techniques, the detection of weak or small targets such as unmanned airplane, skimming over the sea missile, stealth aircraft and warship is usually encountered in the early-warning radar applications. Because of the small radar across section or buried in the strong clutter, the radar echo is weak and the signal-to- noise(SNR)or the signal-clutter-ratio(SCR)is low so that the targets are difficult to be discovered. In order to enlarge the power range of the radar and discover the targets as soon as possible, the detection performance for the weak targets needs to be improved. An effective way to achieve this is to apply the track-before- detect(TBD)algorithm. To detect the targets as quickly as possible and improve the decision proficiency, the sequential detection procedures can be applied. Therefore, it makes great sense to do research on the sequential detection based TBD algorithm.This paper focuses on the detection of the weak targets in early warning radar and the sequential detection based TBD algorithm is mainly discussed. The main contents are as follows:1. Aiming at the likelihood-ratio (LR) based dynamic programming algorithm(DPA), the interplay between the scan frames and system performance is analyzed. Experiment results show that when the strength of the real signals is unknown, in order to get a good trade-off between the early-warning time and detection performance, the chosen of the scan frames is very difficult.2. The FSS and Sequential procedures under the uncoupled (DO) strategy are analyzed. The interplay between two parameters and system performances and the chosen strategies of the parameters are studied. Numerical results have show that if the mismatch of parameters occurs, the sequential detector can diminish the deteriorating speed of detection and estimation performance by adding more scan frames.As the actual SNR is unknown, the number of scan frames for the sequential procedures can be regulated adaptively.3. Aiming at the problem that the estimation performance of the uncoupled sequential procedures is hard to be controlled, the detection and estimation coupled sequential procedures are studied. A detection and position, trajectory estimation coupled(DPTEC)sequential procedure is proposed. By designing a coupled cost function, besides the reliability of the detection and accuracy of the final position estimation, the accuracy of trajectory estimation is also taken in to account, so as to perform the joint detection and estimation. And the detection and estimation performance can be regulated by parameters. Later the detector structure of the procedure is derived and simplified.