Research Of Pulse Compression Radar Signal Intra-pulse Feature Analysis And Processing Technology

Along with the emergence and development of the pulse compression radar, the radar using of pulse compression technology can not only improves the range resolution, but also reduces the probability of interception. As a result, it puts forward the new challenge for the Modern ARM Seeker, which needs the ARM provide necessary intra-pulse modulation feature informatios for the matched filter and interception. Therefore, the intra-pulse modulation feature analysis and processing technology for Pulse Compression Radar signal becomes the critical problem to be solved urgently. The subject is researched deeply and extensively in this dissertation, including the key technology solution aiming at intra-pulse modulation mode recognition, and intra-pulse modulation parameters estimation.The intra-pulse modulation feature analysis and processing system for Pulse-Compression Radar signal is proposed. The Two-Dimensional Spectrum Features's Wave Similarity, its All Parameters Centroid and so on are defined, the modulation feature variable is constructed then. Using the minimum distance principle, the intra-pulse modulation mode recognition method base on Two-Dimensional Spectrum Features's Wave Similarity is proposed. The simulation and anlalysis verifies that, in the SNR and modulation parameters range, this method can achieve nearly 100% accuracy of "modulation category recognition result" among normal, frequency modulation and phase modulation radar category, which is obvious superior to the traditional instantaneous auto-correlation and instantaneous Frequency. It can choose efficient means for the modulation parameters estimation. Then, the improved direction is also discussed.The dual-delay HAF for LFM signal parameters estimation is presented. It can resolve the HAF's contradiction between parameters' estimation precision and unambiguity range, enlarge the unambiguity range. With this improvement, the dual-delay HAF+local ML for LFM signal parameters estimation is proposed. Through the definition of searching range and step for local ML, the dual-delay HAF+local ML not only plays the ML's optimal parameter estimation advantage, but also reduces the computational complexity greatly. In order to improve the real-time performance further, the improved HAF based on Spectrum's Maximum Variance Principle is put forward, combining the dual-delay's demodulation with the local ML's extremum searching idea. The improved HAF can decouple the local ML's two-dimensional joint estimation into twice one-dimensional, and get higher estimation precision than dual-delay HAF. In addition, the generalization of the dual-delay HAF applied to the third-order NLFM parameters estimation and its signal-to-noise ratio loss under multiple correlation operation are also introduced.The PSK parameters' estimation structure based on the baseband signal is proposed, which takes the carrier-frequency estimation as core, using the gradually simplified idea. Firstly, the M-Wavelet-ridge for PSK carrier-frequency estimation and the M-MAT for correcting its deviation are presented. They can acquire not only higher carrier-frequency estimation precision than the traditional and phase differences, but also superior anti-noise performance to phase differences. Using the video synchronous pulse from ARM's microwave front-end and the PSK baseband signal, the structure can reduces the carrier-frequency's influence to the intra-pulse phase mutation detection, and puts forward the PSK symbol rate estimation method based on the two-Scale CWT's detection. Finally, the estimation structure also produces the PSK phase coding sequences estimation method, through estimation of phase codes' difference betweent the intra-pulse subcodes. The simulation and anlalysis verifies that, this strcture can achieve 100% accuracy of BPSK symbol rate estimation when SNR≥3dB, the QPSK symbol rate estimation accuracy can achieve 100% through the correction of phase coding sequences estimation when SNR≥3dB , the BPSK and QPSK phase coding sequences' estimation accuracy can also achieve 100% when SNR≥3dB ,5dB respectively.