Signal Processing And Recognition Technology Of Radar-Communication Integration System

Integrating radar,communication and other equipment to build a multi-platform,multi-sensor information fusion and resource sharing integrated electronic warfare system is the development idea of electronic countermeasures technology in modern warfare and an important topic of current research.At present,the radar-communication integration system has achieved preliminary research results,but it is still in its infancy.Most of the research focuses on signal waveform and system design,but there is relatively little research on integrated signal processing and recognition.With the rapid development of integrated technology,its signal processing and recognition issues are of vital importance for the acquisition of enemy intelligence and war situation assessment.Based on this,combined with the typical radar-communication integrated system,this paper studies the separation algorithm of aliasing signals,modulation recognition of communication signals,radar emitter recognition and integrated target recognition system model.The main work is as follows:(1)For the problem of signal aliasing in radar-communication integration system,the signal separation model is studied,and the fast independent component analysis algorithm is applied to realize the separation of the aliased radar signal and the aliased radar communication integrated signal.In the case where the signal-to-noise ratio is 5 d B,the separated signal obtained by the simulation has more than 96% similarity with the original signal.(2)In the processing of the separated communication signal components,a Stacking-SVM modulation recognition algorithm is proposed to solve the problem that classifiers are high fluctuation and unstable under different signal-to-noise ratios.This algorithm adopts the idea of ensemble learning and improves the single SVM modulation recognition method by using Stacking learning method.By constructing a hierarchical SVM structure model,the algorithm trains the next layer learner with the predicted results of the upper layer learner,and finally obtains more comprehensive essential features of the signal.The simulation results show that the proposed algorithm has high recognition accuracy and better stability under different signal-to-noise ratios.(3)In the processing of the separated radar signal components,an improved radar emitter recognition method is proposed to overcome the shortcomings of radar signal feature extraction and classifier design.In feature extraction,this method extracts texture and shape features of time-frequency image,and combines the correlation features of signal spectrum and instantaneous frequency to form a fusion feature set.In classifier design,this paper applies the method of extracting high-order features based on the GBDT model in the field of radar emitter recognition for the first time and uses this method to extract high-order features of radar signals.Then the extracted features are classified by regularized logistic regression.Finally,the simulation results show that the method proposed in this paper has a higher recognition accuracy under low signal-to-noise ratio.(4)Aiming at the target recognition problem of radar-communication integration system,a system model of radar communication integration target recognition based on multi-sensor fusion is constructed.The model combines the useful information obtained by radar reconnaissance sensors and communication reconnaissance sensors with the results of integrated signal processing to obtain more comprehensive target information,and then uses the target recognition algorithm based on multi-sensor information fusion to obtain more accurate target recognition results.The simulation results show that the uncertainty of identification is greatly reduced and the confidence is increased by using the system model,that is,the system model is effective.