Complex Modulation Radar Signal Modulation Identification And Parameter Estimation Algorithm

In order to enhance the counterreconnaissance capability and surviving capability of the radars in battlefields, as well as to meet the LPI requirement of radar signals, complex-modulated radar signal technology emerging rapidly is widely applied to sorts of new radars and their nettings. Therefore, the detection, recognition, and parameter estimation for complex modulated radar signals has become a hot and difficult point in radar signal interceptation processing.In this dissertation, we focus on several classical complex-modulated radar signals, and emphasize on the recognition and paramter estimation problems. The main contributions are as follows:1. Based on the signal environment of modern radar countermeasure, the models of the Polynomial Phase Signal (PPS), the Sine Frequency Modulation (SFM) signal, the Peseudo-Random Bi-phase Code-LFM (PRBC-LFM) signal and the PRBC-SFM signal are introduced. The properties of time domain and frequency domain for these signals are analyzed.2. For the Nonlinear FM (NLFM) signal, a recognition algorithm based on the 2-order DPT is proposed, which can be used to distinguish the monopulse signal, the LFM signal, the 3-order PPS and the wide-band SFM signal.3. For the FM-PM signal, the Spectral Correlation method is used to recognise the PRBC-LFM signal and the PRBC-SFM signal. Simulation results showed it worked, when SNR > 6dB.4. When the extra-pulse agile signal, the intra-pulse agile signal, the SFM signal, the PPS, the PRBC-LFM signal and the PRBC-SFM signal coexist, the tree discriminator menthod is studied to separate them after feature selection.5. For the PPS, the ML alogrithm, the DPT alogrithm and the IFR alogrithm are discussed, then a converse order judgement algorithm is presented grounded on DPT, which is stabler than the Peleg's one.6. For the SFM signal, two new approaches are adopted to estimating parameters. One is based on Carson Rule, which can directly obtain the parameter information. The other is an indirect menthod, a SFM signal is first modeled as a PPS, by estimating the parameters of the PPS, the needed parameters can be gained finally. Both of the two methods are suitable for NBFM signals as well as WBFM signals, and achieve high accuracy for NBFM signals.7. Square transform is adopted to complete the parameter estimation for FM-PM signals, which is a useful method to separate FM information and PM information. Especially for PRBC-LFM signals, combining with reverse-order correlation method, FM information and PM information can be estimated at the same time.