Radon-Fourier Transform for Radar Target Detection (II): Blind Speed Sidelobe Suppression

This paper gives a detailed performance analysis for the novel radar long-time coherent integration method, i.e., Radon-Fourier transforms (RFT). Some important properties of RFT, e.g., two-dimensional (2D) impulse response, 2D translational invariance, multitarget linear additivity, linear signal-to-noise ratio gain in additive white Gaussian noise (AWGN), as well as the 2D correlation function of transformed AWGN, are derived for continuous and discrete RFT, respectively. However, because of discrete pulse sampling, finite range resolution, and limited integration time, the "blind-speed sidelobes (BSSL)" of discrete RFT may inevitably appear in real applications. Although the BSSL are reduced with the increase of the blind-speed integer, they may still lead to false alarms or loss detections in a real multitarget scenario. Based on the analytic expression derived for BSSL, the causes of BSSL are analyzed and the effective BSSL suppression methods are proposed. Finally, numerical experiments are also provided to demonstrate the effectiveness of the proposed methods.