| Synthetic Aperture Radar (SAR) imaging has been a new developing trend of radar. Based on the principle of synthetic aperture and pulse compression technology, SAR breaks through the limitation of real antenna aperture and realizes 2-D high resolution imaging of distant targets, and it has the characteristics of all-weather, day/night, long range and high-resolution.While SAR imaging has the characteristics of complex arithmetic, large calculation and huge dada, and this requires the system of signal processing has a high performance, large memory, and fast transmission. At present, with the rapid development of the technology of multi-computing and high-speed embedded system interconnect, the study of SAR real-time signal processing has made a great progress.This dissertation is based on a missile-borne SAR item of the cooperation with some institute. The huger data traffic amount and the higher ability of processing are required with the development of the missile-borne SAR real-time imaging technology. Based on the effort of our lab, a SAR real-time signal processing card based on FPGA and DSP was developed.The technique of distributed parallel processing and integrating standard buses with user-defined buses are used in the card concerned. Six DSP are connected in the technology of distributed parallel. They can transmit data in high speed with LINK interfaces, and it is strong in calculations and good in communications. System can be formed by these cards with user-defined buses, and it has powerful arithmetic capability, can achieve real-time high resolution radar imaging.Firstly,this paper presents SAR imaging algorithms.Next,explains the hardware composition of the general purpose signal processing platform, puts forward the technique of distributed parallel processing,compares the popularity bus.Then as an example,expatiates the process of design and debug in detail.At the end,gives the paper's summarize and discuss the new Multicore Digital Signal Processor. |
PDF Full Text Request