Low Probability Of Intercept Signal Identification And Parameter Estimation

In the current complex dense electromagnetic environment, Low Probability of Intercept (LPI) radar signal has been widely in radar and commnucation, due to its low power, wide band, and good anti-interference characteristics, etc. It is difficult to detect and identify the LPI radar signal using traditional interception approaches. And the subject about LPI signal processing has become a difficult and hot issue in the field of electronic reconnaissance. The most important feature of LPI radar signal is reflected in the modulation characteristics, so the modulation identification and parameter estimation of LPI signal become the focus of research in this paper.First, from the view of the radar system, LPI radar theory and the LPI technology have been introduced, 4 typical LPI radar signals (LFM, PSK, FSK and FSK/PSK) were studied, including signal model and the property of LPI. Then, from the view of reconnaissance system, two kinds of signal feature analysis methods (such as the time-frequency analysis and cyclostationary analysis) were analyzed and performance evaluation. Among the time-frequency analysis methods, the author focused on Short-Time Fourier Transform (STFT) and Wavelet Transform (WT) in the application of LPI signal processing. And in the cyclostationary analysis methods, the high-order cyclic cumulants was studied particularly, a fast method for calculating cyclic cumulants was proposed in the paper.There are too many methods for signal modulation identification and analysis today, but on the interception and processing of LPI radar signals, realizing the high probability interception and recognition of LPI signal is the most important goal. In engineering application, the two aspects of algorithm: processing capacity and processing speed, to interception probability are equally important. Therefore, aimed at the interception and identification of LPI radar signals, an effective modulation identification algorithm was proposed in this paper. A recognition description for LPI radar signal, namely phase difference between intra-pulse frames, was extracted quickly through the correlation analysis, combined with support vector machine classifier, automatic inter-class modulation identification of LPI radar signal was realized. Further, realized intra-class modulation recognition among the various identified LPI radar signals, such as within the type of LFM signals between different bandwidths, BPSK and QPSK, 4FSK and 2FSK, etc.In the LPI radar signal paramerter estimation, an algorithm for frequency and symbol rate estimation of LPI radar signal based on STFT time-frequency backbone curve and WT was proposed. Based on the fast cyclic cumulants method, the frequency of LPI signal was estimated accurately. Existing algorithms most are suitable for the analysis and identification of onefold modulation mode, but the most important mode in LPI radar: the hybrid modulation, puts forward a great challenge to these algorithms. In the end of the paper, a synthetical method for multi-parameter blind estimation for hybrid modulated FSK/PSK signal was proposed.