Real time imaging and real time storage of airborne SAR play the important role in civilian filed and military reconnaissance. The advanced countries are trying their best to develop high-resolution, multi-purpose, all digital airborne SAR system.In this paper, based on the research work of SAR real time processor and real time data storage system, we analyze the whole processing of Range-Doppler algorithm. We describe the detail realizing process of parallel digital signal processing platform based on multiple DSP. In the chapter of design and implement of real time storage system, we put forward the design idea based on the computer high performance bus. From the theory and project two aspects, this paper do some valued discussion on the study of SAR real time processor and real time data storage system.Structure methodology is applied to the design of parallel digital signal processing platform. The platform implements the signal processing dataflow of Range-Doppler algorithm such as range compression, clutter lock, azimuth preprocessing and azimuth processing.This paper derives that multiplying 1/2 factor can avoid fixed-point overflow. The improved algorithm based on split radix FFT meet the real time demand of range compression node and remove the effect of pulse compression result for the overflow.This paper analyzes and compares the computation and memory requirement between the two method of estimating Doppler center in time domain and frequency domain. The result is that time domain method is effective. And derive the fast flow to implement the estimation of Doppler center on the platform.This paper implements azimuth preprocessing before comer turn memory step based on the DSP platform. The linear-phase FIR filter is designed by using the Parks-McClellan algorithm and can vary with the Doppler center.By analyzing the realization of range migration, we conclude that the pipeline structure is effective to implement azimuth processing.The memory mapping technology and multithreaded I/O operation of SCSI processor is applied to the design of real time storage system.