Abstract: If the suppress jamming pour into the mainlobe of radar antenna, it would severely degenerate the performance of radar. The common ECCM measures of sidelobe have no effect. But the performance of the technique of anti-mainlobe-jamming based on blind source separation algorithm would deteriorate in the condition of low signal to noise ratio. In this paper, we present the technique of anti-mainlobe-jamming based on BSS with the fractional Fourier transform de-noise. The mixed interference and target signal is separated by the classical BSS algorithm in the first of all. Then the signal separated which is a linear frequency modulation signal is dealt with FRFT and filtered to cut away the almost energy of noise in FRFT domain. The target signal is recovered by the FRFT opposite transform at last. The experiment of simulation indicates that it is a superior improvement for the peak SNR after pulse compression for the new algorithm than the process of BSS. Because the detection performance of the pulse compression radar is enhanced preferably, the new algorithm can be applied in practice.