By the use of the principle of compressed sensing (CS) theory which is based on the sparsity or compressibility of signals, a new sparse representation for azimuth is proposed through the analysis of synthetic aperture radar echo signal. On this basis, a SAR echo signal processing method is obtained and a 2D SAR imaging algorithm which achieve a fire-new data collection and coding of signal is established. The use of this algorithm not only suppresses the side lobe effectively but also enhances the resolution of imaging target to a certain extent, and simultaneously needs less amounts of data for imaging, which provides an efficient path of decreasing the data rate of high resolution synthetic aperture radar. The feasibility and validity of the method proposed in this paper are tested through processing both simulated data and real radar data.