SAR Imaging Algorithm And GPU Acceleration Research Based On Backprojection

Synthetic aperture radar(SAR) is a high resolution microwave imaging system with all weather, day and night capability. The characteristic of high resolution makes it play an irreplaceable role in the military and civil fields. With the development of SAR imaging technology, all kinds of high resolution imaging algorithms arise. However, tremendous computation amount caused by high resolution becomes the bottleneck of some imaging algorithms in practical application. And among them, BP algorithm is a typical time-domain imaging algorithm. Compared with the traditional SAR imaging algorithms, it is simple in principle and has no theoretical approximation. Therefore, BP algorithm has better imaging result and more practical study value.Based on the above backgrounds, the main work of this paper are as follows:(1) We analyze the BP and FBP imaging algorithms, including their imaging models, realization principles and computation amount analysis. The BP algorithm can’t be used widely because of tremendous computation. This paper studies a FBP imaging algorithm based on BP algorithm. And the experimental result shows that the imaging qualities of FBP algorithm and BP algorithm are almost the same. And FBP algorithm decreases the computation amount to a certain degree at the algorithm level.(2) We also sudy the motion error and motion compensation method of BP algorithm in the case of non ideal trajectory. The error source of BP algorithm is ananlyzed and the motion error model is establised. By Introducing the motion compensation technology of BP algorithm, we propose two phase compensation algorithms based on contrast maximization criterion and subaperture-contrast maximization criterion, respectively. And the final imaging results of the measured data verify that the two motion compensation algorithms can be effectively applied to respectively short aperture and long aperture.(3) We introduce the GPU programming technology based on CUDA, and Analyze the inherent parallelism of BP algorithm. A scheme suitable for GPU acceleration is also proposed. Compared with the SAR raw data amount, GPU memory is not enough. Then GPU acceleration optimization is proposed based on stream technology. The result of the measured data shows the effectiveness and feasibility of the scheme, and the averge speed of optimized algorithm improves 78.8 percent.