A new radar TBD method based on RD-CTS

Track-before-detect (TBD) is a useful strategy for weak target detection that manages to integrate target energy along candidate trajectories and test whether the summation is large enough to declare a target. Several methods have been proposed to apply TBD in radar weak target detection, but none of them could make profitable use of the inherent characteristic of radar data. A simple and time saving radar TBD method is proposed, which is unique in an innovative Range-Doppler Candidate trajectories structure (RD-CTS). A number of candidate trajectories are hypothesized off-line, taking advantage of the correlation of the Doppler and range of each "root" image cell. Measurements along possible target trajectories are then integrated and a target is declared if the measurement sum, or merit function, exceeds a threshold. The CFAR (Constant false alarm rate) detection and false alarm probability determination are analyzed as well. Under a practical set of radar parameters, good detection performances are presented that at about 100% detection probability, 7.5dB SNR gain is achieved through 15-frame non-coherent integration.