Recognition Of Deception Jamming Based On Mutual Information Entropy Under Multi-radar System

Electronic countermeasures(ECM)technology is widely used in modern electronic warfare.With the increasing complexity of electronic frequency spectral density,more and more interferences affect the performance of radar,and the living environment of radar has been challenged as never before.Radar deception jamming is one of the most important interference techniques in radar electronic warfare.Especially with the rapid development of digital radio frequency memory(DRFM)technology in recent years,the deception jamming signal can be generated as the coherent replica of the intercepted radar target,which makes the discrimination of the radar false target more difficult,and brings tough challenges to the normal function of the radar.Therefore,it is necessary to take anti-jamming measures to extract the statistical characteristic parameters of deception jamming signal,so as to identify the jamming.How to choose the effective anti-jamming method for radar system is a hot research field currently.In this thesis,considering from the perspective of the information theory,the mutual information entropy and the generalized correlation function are extracted as classification features,combining with the idea of fusion algorithm,the detection and identification of deception jamming signal are realized in multi-radar system.The main research work of this thesis is as follows:1.Firstly,the research background and development status of deception jamming detection and identification at home and abroad are introduced.Then,the current common radar active deception jamming patterns as well as the radar interference technology based on DRFM are introduced and the corresponding signal models are established.The mechanism and performance of different deception jamming signals are described,which lays a theoretical foundation for the feature extraction of active deception jamming signal.2.Algorithms for identifying deception jamming based on mutual information entropy and generalized correlation function are studied.Firstly,for the defect that the Pearson correlation coefficient cannot describe nonlinear relationship well,mutual information entropy is introduced into the detection and recognition of the jamming.A deception jamming recognition algorithm based on mutual information is proposed.The simulation results verify the feasibility of the proposed algorithm.Compared with the method based on high-order spectral diagonal slice variance,the results show that the proposed algorithm has better recognition performance,the average recognition probability is increased by 2.72%.Then,a deception jamming recognition algorithm based on generalized correlation function is proposed.The algorithm uses the equal spacing algorithm to estimate the mutual information entropy,and then the mutual information entropy is normalized,which further improves the recognition performance and the average recognition probability is increased by 2.53%.3.The multi-radar system deception jamming identification method based on mutual information entropy and generalized correlation function are studied.Firstly,the fundamentals of information fusion in multi-radar system is introduced.Then,the mutual information entropy and the generalized correlation function are applied as the classification features in each classifier.According to different fusion rules,the data information in each sub-classifier is correlated and unified processing is performed in the fusion center,information fusion is realized from the perspective of feature layer and the decision layer to obtain the decision result.The simulation results verify the effectiveness of the algorithm.Compared with the monostatic radar,the correct recognition rate of the deception jamming by the three decision layer fusion algorithms:Bayesian estimation method,Bagging algorithm and DS evidence theory are increased by 9.18%?11.04%?17.43%on average as well as the two feature layer fusion algorithms:weighted average method,principal component analysis method are improved by 16.59%?21.31%when selecting mutual information entropy as the classification feature,and when the generalized correlation function is used as the classification feature,the average recognition rate of the three decision fusion algorithms are increased by 9.85%?12.45%?17.03%,and the recognition rate of the two feature layer fusion algorithm are increased by 15.19%?22.64%,which proves the effectiveness of the fusion algorithm.