Detection And Analysis Of Low Intercept Probability Radar Signal Based On Compression Sensing

With the development of surveillance and counter-surveillance techniques, radar performance requirements are not only limited with finding the target, but ensured that the receiver will not be intercepted signal reconnaissance. LPI radar is born in order to complete the above functions. With the LPI waveform bandwidth is increasing, Nyquist sampling theorem based processing algorithms to achieve more and more difficult to intercept the signal data acquisition. To solve this problem, this paper proposes a method using CS theory to complete the signal sampling, reconstruction and detection analysis.Firstly this paper analyzes related characteristics and ambiguity function of the three typical LPI signal, P4 code, costas and step frequency modulated pulse signal. In response to these signals, considering of the principle in inter pulse FM and pulse phase modulation, the paper puts forward a new compound modulation signal model,including costas combination with p4 and step frequency modulated pulse signal combination with p4. The compound modulation signals have big time-bandwidth product features, at the same time, and improves the performance of the original signal and ambiguity function is more close to the Pushpin-Style. The composite modulation signals make more complex and the bandwidth is obtained to expanding, which is difficult to be identified by aircraft.Secondly, analysis of intercepted signals. At present, with the direction of development of big time-bandwidth product signals and conventional electronic reconnaissance aircraft difficult to detect, this paper puts forward the analysis method of the signal detection, which is based on compressed sensing. The method can complete sampling of a signal in the low rate and reconstruction and analysis of signals In this paper, LPI waveforms of typical and complex modulation are be sampled and reconstructed. Then, analysis the reconstructed signal by using the time frequency distribution of CWD.Finally, the research is finding that the noise has a great influence on signal reconstruction to affect the time frequency analysis of signal reconstruction.In order to solve this problem, the paper puts forward another method that with no need to reconstruct the signal, it directly samples signals to obtain the time-frequency analysis graph. The method can reduce the influence of noise on signal analysis.