Research On Tbd Technique For Radar Weak Target

The properties of modern wars, such as omni-direction, multi-layer and great-depthmake the covertness of targets become better and better, which diminishes the power ofradars in target detection greatly. Consequently, weak target detection is a crucialproblem that is starving for being resolved in the field of radar signal processing.Track-before-detect (TBD) algorithm is proposed for this task.TBD algorithms based on dynamic programming and particle filters are mainlydiscussed in this article. Aiming at radar weak single-target with different characteristics,three improved detection methods are proposed. For a weak-maneuvering target, anapproach using coherent integration and TBD based on dynamic programming isadopted. Moreover, directional weights and a technique of two step detection thresholdsare introduced to enhance detection performance. For a short-time strong-maneuveringtarget, a new detection algorithm based on particle filter is proposed, in whichhigh-precision particles production mechanism and effective partition particles areutilized to improve the diversity of particles. For a strong-maneuvering target, a newTBD algorithm based on particle filter and dynamic programming is proposed. In thenew algorithm, the weights of particles in turn are taken as the evaluation function ofdynamic programming, thus scans of data being integrated. And this strategy can makethe algorithm be parallel. The results verify that the new methods can enhance thedetection performance and efficiency effectively compared to several detectionalgorithms proposed recently. At last, an approach based on time-frequency analysis andTBD is utilized to detect radar weak multi-target based on different signal models. Theresults demonstrate that the new method can enhance the time-frequency collection ofsignal effectively.