Research And Implementation On Coherent Integration Of The TV Based Passive Radar On GPU

The passive radar which makes use of broadcast, communication, television andothernon-cooperative illuminators can be configured to bistatic/multistatic system for targetdetection andlocation. This radar has the merit of resisting the radar’s four major threats,and hasreceived considerable interests.As common civiltransmitters, digital TV signalsused in passive radar have very important meaning. Coherent integration algorithmis animportant technology to improve detection range and the ability to detect weak targets fordigital TV based passive radar. The coherentintegration algorithm andalgorithmimplementation on GPUforpassive radarusing digital TV terrestrial broadcastingsignalsisresearched in this paper, based on the supports of the National Natural ScienceFoundation of China. The research results of this paper are fundamental for theimplementation of the passive radar using digital TV terrestrial broadcasting signals andhave theoretical and practical value. The main innovations and contributions of thisdissertation are summarized as follows:Firstly, we introduce the background and significance from the perspective of passiveradar system,coherent integration algorithmand implementation on GPU.Secondly, the geometric model of passive radar system is analyzed, and echo signalmodel of target is derived.The coherent integration algorithm based on calculationof thecross ambiguity function is researched.The coherent integration algorithm to achievesegmentation method based on pulse compressionis proposed.Again,the reason why range migration is produced on uniform motion of the target isanalyzed. Then the algorithms of long time integration with migration compensation basedon keystone are proposed. Andthe three approaches of keystone transformareestablished.The algorithms based on low pass filter and decimatorcombined chirp-ztransformwith migration compensation isproposed.Finally,Nvidia’s GTX460chips are studied in detail and its data storage architecture,computational model, programming architecture are analyzed. The algorithms above workto achieve and optimize.