| With the SAR (Synthetic Aperture Radar) system developing to high resolution and wide band, the SAR data is become larger. Processing the huge amounts of data quickly and efficiently has become the key of SAR imaging and information extracting in remote sensing. The rational design of parallel processing hardware system using characteristics of the SAR ground data processing becomes the core to solve the problem of massive data processing.Parallel processing system is one of the most effective methods to improve computing speed problem of large-scale data. Through combining the characteristics of SAR data processing and the existing computing technology to design the parallel processing system reasonably, rapid processing of massive data can be achieved. As a result, related remote sensing information and images also can be obtained accurately, which is the foundation of SAR application. In this thesis, with parallel processing hardware system of SAR data as the research object, the concepts and calculation methods of the main systematic design index are given. Combined with the characteristics of SAR data processing, cluster parallel processing system, blade parallel processing system and cluster GPU (Graphics Processing Unit) parallel processing system, parallel processing hardware system for SAR data processing is discussed in detail. This analysis provides a reference for SAR parallel processing hardware system design.According to the SAR data processing characteristics, this thesis gives the main design index of parallel processing. At the same time, combined with the existing parallel processing hardware technology, the design and implementation of parallel hardware system are studied. Moreover, three kinds of different ways of the parallel processing hardware system are discussed in detail. They can provide the reference for SAR data processing and reasonable parallel processing hardware platform building, and create more favorable conditions for SAR applications. |
PDF Full Text Request