TBD Algorithm For High Frequency Surface Wave Radar Based On Polarization Information

Through the research on target detection for high frequency surface wave radar, we have found that there are a big amount of interference in the HF radar echo data, such as ionosphere clutter, sea clutter, radio frequency interference and so on. The energy of these interferences are very strong and these interferences are distributed in a wide range. All of these cause the dim target echo to be covered by complex clutter and noise. How to detect and track the target effectively is the key point of this paper. Track-before-detect technology is a very effective method for weak target detection and tracking. In recent years, more and more TBD algorithms have been applied to the field of radar target detection and tracking. Polarization information is an important characteristic of electromagnetic wave as a vector wave. Target detection using polarization information is also a rapidly developing field. Combined the polarization information with TBD technology and applied them to the HFSWR target detection and tracking field is the method that this paper focused on. The contents of this paper are as follows:At first, the polarization state of target and clutter witch in HFSWR target echo data is analyzed and its distribution and statistical laws are given. This is the basis of using polarization information to detect target.Then we research the basic TBD technology. We focus on dynamic programming algorithm and give its principle and simulation results to prove that the application of dynamic programming algorithm is feasibility.In the third part, we analysis the limitations of dynamic programming algorithm and give an improved algorithm. By using the prior knowledge of the target and the polarization characteristic of the clutter, the new method is used to improve the dynamic programming algorithm.At last, the dynamic programming algorithm is extended to the background of multi object detection and tracking, and the corresponding dynamic programming algorithm is proposed to realize the tracking of multiple targets.