Study On Design Of Real Time Embedded SAR Imaging System

Synthetic Aperture Radar (SAR) is an important tool for the collection of all-weather image data. Furthermore, these data can be made at extremely high resolution by signal processing. The SAR achieves the high resolution of radar image by the technology of pulse-compression in range and by the principle of synthetic aperture in azimuth. It is widely used in the fields of disaster forecasting, military inspecting, terrain mapping and resources investigating, etc.Aircraft SAR imaging systems are fixed on the plane, in which SAR signals can be collected, processed, stored and imaged in real-time. These systems are always consisted of multiple embedded signal processors, and are the typical example of embedded systems applied on high-performance real-time computing. Different from massively parallel processing systems, data are managed and processed as "Data Flow" in embedded systems. The architecture of these systems is distributed and heterogeneous.In order to study the design method of aircraft SAR real-time embedded imaging systems, the characters of high-performance real-time embedded systems are analyzed in this paper.1. Model analysis is important to design of embedded systems. This paper provides the Enhanced PGM Model, which is a data flow graph model and suited to analyze distributed heterogeneous real-time embedded systems.Firstly, based on the analysis of queues in the model, three patterns are provided to describe the throughput characters of embedded systems. These patterns include sliding window, data increasing mode, and data decreasing mode. Secondly, embedded systems are categorized into three parts by the real-time characters: standard model, hard throughput model and hard end-to-end latency model. At last, a novel method is proposed to use Enhanced PGM to compute throughput and end-to-...