Research On Real Time Pulse Train Deinterleaving For Radar Intercept System

Working silently and covering larger region than Radar,Passive radar intercept systems have superior survival ability in wars. It's widely used in many weapon systems. In high dense complex environment,traditional pulse trains deinterleaving methods can't fulfill the real time processing requirements.In this dissertation,several technology problems of pulse trains deintrleaving algorithms are dealt with,they are presorting techniques based on coherent processor,probabilistic neural network deinterleavers,adaptive data association methods for pulse trains analysis and deinterleaving,signal processor designing issues. The research is focused on real time processing.The coherent processor is a crucial technique for real time presorting. Two new methods of designing coherent processor are addressed in Chapter 3. These coherent processors can fulfill arbitrary complex parameters. Chapter 4 puts forward two deinterleaving methods based on probabilistic neural network (PNN). The self-organization PNN deinterleaver is suitable for unknown emitters,while the RBPNN deinterleaver is suitable for known emitters. By employing the state-space model for pulse trains,kalman filter and probabilistic data association,the thesis develops a deinterleaving algorithm for tracking mode in Chapter 5. The Interacting Multiple Models (IMM) approach is implemented to simultaneously estimate the pulse train PRIs of multiple emitters.The radar intercept system signal processor designing issues are also discussed in Chapter 6.An architecture of signal processor is brought forward,while several examples support it. The advantages of real time operating system (RTOS) are illustrated,and the approach to develop signal processor software infrastructure is probed into. At the end of thesis,the parellel design problem is summarized.