Frequency Analysis Of LPI Radar Signals

LPI (Low Probability of Intercept, LPI) radar has low peak power, large time-bandwidth product and other outstanding advantages, which leading the traditional intercept receiver be difficult to detect LPI radar signal. And the modulations of LPI radar signal are more complex, which made the traditional signal parameter estimating algorithms have a serious decline in performance or even failure. Begin with the technical principles of LPI,this paper proceed on a number of classic LPI radar signal performance analysis, combined with the non-stationary characteristics of LPI signal, this paper also made a detailed study of the performance of using the time-frequency distribution algorithm to detect and estimate the parameter, and had a optimization of the algorithm. The main contents include:Firstly, the principle and technology of LPI has been studied, and the paper has a detailed discussion of LPI radar combined with the concept of intercept factor. Providing signal model and basis data for follow-up studies, a number of typical LPI radar signals have been studied and simulated in terms of time-domain, frequency-domain and fuzzy function.Secondly, the WVD and PWVD have been studied, the cross-term representation of WVD has been derived and the time-frequency aggregation also has been analysised. Using PWVD algorithm to simulate the LPI signals in different noise environment, the results show that the algorithm has a good time-frequency aggregation, but there have been serious cross-term and reduced the performance of the algorithm greatly.Thirdly, for WVD with large cross-terms, the kernel of WVD has been improved combined with the concept of Cohen. The principle of CWD and the representation of cross-term have been discussed in detail. Using CWD, the LPI signals have been analysised and simulated. The results of simulation indicated that although the cross-term has been combated, its time-frequency clustering has been deteriorated dramatically.Finally, for large cross-terms or bad time-frequency aggregation, the method of using the source entropy to improve the performance of CWD. Principle of the optimization algorithm has been discussed at the first, and specific implementation steps of the algorithm have been given. Then, the simulation of LPI signals have been taken by using the optimized CWD and the results indicates that the performance of optimized CWD was more excellent.