Research On CBP Algorithm And Implementation On Parallel DSP System

Synthetic Aperture Radar(SAR) is a kind of imaging radar with fine resolution capability and the possibility of day/night operation in all weather, which is of great importance to the modernization of national defence technology and national economy development. SAR have three kinds of common imaging mode: Stripmap Mode, Spotlight Mode and Scan Mode .As an important imaging mode of SAR, Spotlight Mode SAR can obtain high azimuth resolution that might be hardly obtained by the conventional stripmap SAR. This dissertation studies the convolution backprojection(CBP) algorithm which is used in the image formation of the Spotlight Mode SAR through the processing of the simulated data and live data that was converted from Stripmap Mode radar. Based on the excellent parallel character of CBP, parallel processing system consisted of multi-processor, Darlingtong development board, is used to implement CBP algorithm with live data. The main contents of this dissertation are as follows: Chapter 1 introduces the development and research situation of SAR and Spotlight Mode SAR in detail. Finally, the main research contents of this paper are summarized. Chapter 2 introduces the image formation theory of SAR and analyzes CBP algorithm. Chapter 3 presents the interior architecture of SHARC processor and the technologies on parallel processing system consisted of multi-processor, and analyzes and discusses parallel algorithm theory. Chapter 4 sets down the hardware realization methods on parallel developing system, points out some problems to be noticed, gives the improved methods and presents the processing results. In the end, the paper summarizes the whole work, and points out tasks to study and the way to continue researching.