Study On Realization Of Dim Moving Target Detection And Tracking For Radar

Dim moving targets’ detection and tracking have been hot topics in warning radar.With the rapid development of stealth technology, the performance of detection andtracking always suffer a challenge. As we know, weak targets are difficult to be detectedbecause of its low echo signal to noise ratio. We usually increase the signal to noiseratio via signal processing, and an effective method is long-term integration to theenergy of its echo signal which can be divided into coherent integration of single-frameand non-coherent integration of multi-frame.Based on different characteristics of dim moving targets, two modes of theaccumulation of energy are studied and dynamic programming (DP)-basedtrack-before-detect (TBD) algorithm is used for engineering realization in this thesis.First, for targets of high mobility, long-term coherent integration in single-frame isanalyzed. Keystone transform is discussed for solving range migration of target echocaused by speed when long-term integration. And the fractional Fourier transform(FRFT) method is discussed for solving the problem of spectral diffusion caused byacceleration. Then, the method of non-coherent integration in multi-frame is analyzedfor targets of weak mobility and mainly discussed the TBD method based on DP. Forthe limitation of conventional method, an improved method is proposed. In the newmethod a value-transfer function is constructed which just uses the variable of range andspeed, estimated target’s track, then further confirmed track by integrating scans of dataalong the track. The results verify that the improved method can enhance theperformance of detection and tracking compared to algorithms proposed recently. At last,radar signal processing engineering implementation technology is introduced combinedwith the improved method of DP-TBD, which further validates the effectiveness of themethod.