A SAR Raw Data Compression Algorithm Based On Imaging

Synthetic Aperture Radar (SAR) is one of the most important remote sensing means for both military and civil purposes, not only because of its ability of working under all the weather and all the days, but also the characteristic of long range and high-resolution. As the development of digital signal processing technology, SAR imaging has become essential for efficient SAR systems. In the mean time, the conflict between the large data rate of the raw data and downlink speed is becoming more and more obvious. An efficient method to solve this problem is using the raw data compression algorithm.In this paper, SAR imaging technology is presented at first, including the imaging theories and characteristic of SAR signal. Range Doppler algorithm (RDA) and Chirp scaling algorithm (CSA), which are two kinds of classical imaging algorithms, are analyzed in detail through point targets simulation. Range cell migration correction (RCMC) is indispensable of SAR imaging. Other than those algorithms such as RDA implement interpolation in range-Doppler domain, a RCMC that implement in azimuth-time domain is proposed. Then, the theories of data compression and quantization are introduced, and the statistical properties of SAR raw data are analyzed and computed using the simulation data. Block adaptive quantization (BAQ) algorithm is discussed, which is the first SAR raw data compression algorithm.Compare to the real and imaginary data of echo signal, the phase component that contains Doppler information of targets worth more research. The amplitude data obeys Rayleigh distribution, while the phase data obeys Uniform distribution. Thus, an algorithm using optimization quantizer and uniform quantizer to compress amplitude and phase components respectively is adopted in this paper. Experiments prove its feasibility, and show that this algorithm is superior to the BAQ algorithm under the same compression ratio, while reserving phase information better.