The active sonar target echo is seriously interfered by the marine environment noise. The long-range detection echo signal is weak, and the target azimuth estimation accuracy is low. Moreover, the traditional matrix spatial spectrum estimation method needs to satisfy the Nyquist sampling rate and the data acquisition amount is large. Based on the spatial sparseness of the signal and the echo signal bright spot model, the spatial azimuth estimation technique based on compressed sensing is studied. In the case of limited physical aperture of the receiving matrix, the linearaperture(LP) virtual array method is used to improve the aperture size of the array, and the compressed echo(CS)algorithm is used to reconstruct and recover the target echo signal. The data is simulated and analyzed. The resolution of the spatial spectrum estimation of the array is improved, and the noise interference is effectively suppressed.