Multi-fold high-order cumulants based CFAR detector for radar weak target detection

Abstract

As an important detection means, radar is widely used in target monitoring field. However, due to the small radar cross section (RCS) of stealthy or small target and the complex detection environment, the target echo signal is usually submerged in the environmental clutter. Then the target detection based on Pulse-Doppler radar shows the characteristics of weak signal detection. In this paper, the multi-fold high-order cumulant (HOC) on the radar range-Doppler (R-D) map is proposed to elevate the detection performance of the weak target signal, which can effectively eliminate the noise. Then the processed signal is sent to the CFAR detector to estimate the targets' range and velocity. Although HOC has been used in radar signal detection, the detection algorithm was mainly done on the processed signal after one-fold HOC. We propose a construction method of multi-fold HOC using a set of delay vectors to further eliminate residual noise after one-fold HOC processing. Our proposed multi-fold HOC can be seen as a data preprocessing method, which can also be applied to other detection algorithm to further improve its detection performance. Simulation results show that the proposed multi-fold HOC based CFAR algorithm can greatly reduce SNR required for signal detection, and has an advantage in radar weak signal detection.