Radar compressed sampling with longtime pre-coherent integration via RFT

Compressed sampling is generally used in ultra-wideband radar, in which the moving targets' radial velocity would easily cause the across range unit (ARU) effect due to the high refined range resolution. Meanwhile, the noise-folding effect caused by under-sampling reduces the signal-to-noise ratio (SNR) of compressed sampled signals apparently. Therefore, traditional pre-coherent integration via discrete Fourier transform (DFT) are very likely to be unable to provide enough SNR gain for reliable reconstruction. In this paper we propose a novel radar compressed sampling system with longtime pre-coherent integration via Radon Fourier transform (RFT). The compressed sampled signals are firstly decompressed, then integrated by RFT filter banks, and finally reconstructed on each velocity bin respectively. The integration time of the proposed system is no longer restricted by the velocity of targets and range resolution as in the traditional pre-coherent integration via DFT, but could be greatly extended to satisfy the actual SNR requirements of system. Numerical simulation results show that the rate of successful reconstruction of provided system with high targets velocity is significantly improved, compared with the previous method using DFT integration.