Research And Engineering Realization Of Radar Signal Processing Algorithm Based On TigerSHARC

Monopulse Radar is a kind of advanced radar using coherent processing. In this paper, we design and realize a radar signal processor based on TigerSHARC DSP system. The signal processor can acquire and track the target in appointed airspace range and offer target parameter in real-time, such as relative distance, relative velocity, azimuth angle, elevation angle and angular velocity.The paper is involved with algorithms realization of pulse accumulation, constant false alarm rate(CFAR) Detection, parameters estimation and target tacking.First, The paper introduce hardware resources of ADSP TS101S, and introduce the interruption of resources, link port and DMA transmission emphatically. The Integrated Development Environment for ADSP chip is also summarized. The More crucial part of mixed-language programming and the LDF file are described detailedly.Then, the concrete realization of the monopulse radar signal processor is introduced. The paper research and realize the algorithms, such as data reception, pretreatment, pulse accumulation, target detection, and target evaluation. In the research of pulse accumulation algorithm, we propose two kinds of schemes which are DFT coherent integration algorithm and FFT noncoherent Integration algorithm based on the requirement of DSP resource and Real-time Property. The latter is adopted comparing the two methods with respect to their advantages and disadvantages. In the research of target detection algorithm, we propose the slide window CFAR detection algorithm based on outlier removal. we estimate distance and velocity parameter and track the target.Finally, we do research in tracking algorithms for single target. The paper introduce three kinds ofα-βfiltering algorithms which are the constant gainα-βfiltering algorithm, an adaptiveα-βfiltering algorithm in Spherical Coordinates and an adaptiveα-βfilter algorithm in mixed coordinate system. And do a performance comparison among the three algorithms through Monte Carlo simulation. Becauseα-βfiltering is based on distance information, not using the accurate velocity information, so we propose a real-time tracking algorithm based on distance and velocity information. The Monte Carlo simulation results show that the new algorithm has a better performance thanα-βfiltering and is a little inferior to Kalman filtering. Moreover, the new filtering algorithm matches real-time request.