With the rapid development of SAR missions

on-board SAR image processing is required. SAR could obtain much higher resolution images than ordinary radar

and it is very necessary to compress data before transmission

since the generated SAR data volume is much higher than the data bandwidth of the satellite downlink. In this paper

an efficient compression algorithm with accurate rate control is proposed based on SAR complex image characteristics. The algorithm applies an adaptive weighted quantizer on 2D-FFT coefficients

and then uses entropy encoder to code the quantized coefficients. Based on the es-tablished R-Q and MPE-Q statistical models

an accurate rate control is developed. We measure the compression performance with PSNR and MSSIM （Mean Structural Similarity） of the magnitude image and MPE （Mean Phase Error）. With four SAR complex images compressed at different ratios

the compression performance is analyzed and compared with other algorithms including JPEG2000

H.264

etc. Experimental results show that the proposed algorithm outperforms other algorithms

and the rate control obtains high accuracy.