Study On The Autonomous Processing Methods Of Hybrid Modulated Radar Signals In Electronic Reconnaissance

Modern radar systems are gradually developing their abilities in environment cogition and adaptive detection,and the emitted signals of them have the ability to modulate and agile parameters in multiple domains.More and more new and hybrid modulated radar threat signals are emerging in modern electromagnetic environment,which brings a great challenge for the signal processing in radar reconnaissance receivers.For the receiving and processing of these kinds of radar threat signals,such as the hybrid modulated signals,it has become an urgent requirment to improve the autonomous processing ability of the intercept receivers.In this dissertation,several autonomous processing frameworks are proposed based on the aspects of algorithm modules,signal processing structures and cooperation between different systems.Some algorithms,such as autonomous modulation recognition methods for hybrid modulated radar signals,autonomous frequency estimation of a single frequency pulse,and fast feature extraction of five kinds of linear frequency modulation(LFM)derived polyphase coded radar signals,are proposed corresponding to the level of algorithm modules.The parameter estimation performance of the the hybrid modulated signal which combines the pseudocodes and LFM is also analyzed.These researches are able to provide useful ideas for radar interception receivers and the other related electronic warfare technologies.The main contributions and innovative results of this dissertation are as follows:1)Several feasible autonomous processing frameworks in radar reconnaissance are studied and discussed based on some signal processing aspects.On the aspect of algorithm modules,the autonomous frameworks based on parallel processing and information fusion,and frameworks based on the iteratively approximating method are proposed.Detailed framerworkds are discussed to analyze the intrapulse and interpulse features respectively based on signal-to-noise ratio(SNR)values,modulation components analysis,multi-channel modulation parameter estimation,parameters with different confidence levels and parameters with different reliability coefficients.On the signal processing process aspect,a parallel and feedback framework which needs interactions between different modules,and a deep learning structures for the electromagnetic environment situations are proposed.The latter structure emphasizes on the situation awareness of the electromagnetic environment and its describing,backtracking,tracking and prediction.Thus frameworks including the parallel processing of the known and unknown threats and signal sorting using the backtracking of pulse sequences are designed.The allocation frameworks of resources based on the threat levels and the number of simulantaneous arriving signals are also given.On the aspect of cooperations between friendly systems,the reconnaissance system is considered to improve the anti-jamming capability for a friendly radar system and improve the effectiveness for an electronic countermeasure systems.2)An autonomous recognition method is proposed for the new and unknown radar signals incluing the hybrid modulated ones,based on the proposed modulation component analysis method which uses the properties in the instantaneous frequency rate(IFR)curves.Firstly,the radar intentional modulation on pulses(IMOPs)are classified into three families including the continuous frequency modulation(CFM),discrete frequency coding(DFC)and discrete phase coding(DPC)families.And the DFC and DPC families could be divided further into the basic and hybrid IMOP types.According to the theory of generalized representation of phase derivatives(GRPD)and the measured SNR values,the IFR curves are derived using an adaptive sliding window.Then the unknown modulations are recognized without a priori knowledge according to the IFR curves,and this can be used as pre-recognition in the existing recognition architecture without performance loss for the other signals,Simulations and comparison demonstrate the effectiveness of the proposed autonomous framework and algorithm.3)A fast parameter estimation algorithm based on the reliability coefficients is proposed for the single frequency pulse.The frequency can be estimated autonomously based on three existing algorithms whose estimating errors are within the frequency resolution of the discrete Fourier transform(DFT).This method not only can have a performance close to the Cramer-Rao lower bound(CRLB)for adequate SNR values,but also can provide reliability evaluation for each output.4)For the five kinds of LFM derived polyphaser coded signals,two parameter estimation method are proposed.The first estimation method is based on a parallel processing framework and the instantaneous autocorrelations with different delays are applied.A second method that can simutanuously estimate the center frequency and code rate of polyphaser coded signals using only the iterations of DFT coefficients is also proposed.Theoretical analysis and simulation results demonstrate that the proposed autonomous frameworks and methods are effective.5)The modified CRLB(MCRLB)for the parameter estimation of the hybrid modulated radar signal combinng pseudo-code and LFM in white Gaussian noise is deduced and the performance of the step-by-step estimation method is also studied.The corresponding results could provide a theorectical evaluation tool for the estimation algorithms of this signal and are meaningful for the other similar hybrid modulated radar signals.