Interference Detection Technology Research Of Air-Ground Intercom System Of The Civil Aviation Based On Feature Representation

In the increasingly complex modern communication environment,it is especially important that the technical problem called interference detection need to be solved firstly for the interference problem in wireless communication.It is positive for guaranteeing the quality of communication to effectively detect the existence of interference and make the relevant treatment,whether our communication equipment will face the malicious interference from enemy like suppression jamming and deception jamming in the modern electronic information warfare,or civilian communication will face the intermodulation interference in the internal communication system and the interference in the external communication system.Firstly,this paper illustrates the background of the interference problem in the air-ground intercom system of the civil aviation,artificial neural network and deep learning,and analyzes the research status of interference detection in this system.Then this paper researches the feature extraction of the signal,and mainly analyzes the basic signal feature parameters like the time-frequency domain feature and the constellation diagram and other feature parameters such as approximate entropy,circular statistics and higher-order cumulant.The feature extraction of interference detection and signal recognition is based on these characteristics.Secondly,this paper proposes the interference detection method with the BP neural network model using multiple characteristic parameters in the interference detection problem.This method uses the BP neural network to map the relation of multiple characteristic parameters by nonlinear space,and combines with the network output layer to avoid the selection of threshold in traditional interference detection method.This paper simulates the performance of the interference detection with this method using the fixed AM signal in the air-ground intercom system of the civil aviation and the interference signal of the FM signal in the FM radio as the signal analysis.The result of the simulation analysis shows that the method can realize the self-independence extraction of multiple characteristic parameters in the interference detection,and reduce artificial error in the selection of threshold.In addition,the method improves the accuracy of the interference detection.Then for the artificial extraction of the characteristic parameter in the interference detection,this paper puts forward an autonomous and intelligent interference detection method using the feature representation of the deeper network structure,which is different from the traditional interference detection.This method directly regardes the time sequence of the signal as input sample of the network with the deep learning network model of the deep network structure,and uses nonlinear transformation of the network layers to extract the characteristic parameter of signal.And then this method regardes trained network as the characteristic parameters of signal.Using the simulation environment of the interference detection method based on BP neural network,this paper simulates the neural network model with the built convolutional neural networks.The result of the simulation analysis shows that using the feature representation of the deeper network structure to solve the interference detection problem is feasible.This method combines with the signal parameter extraction and interference detection in deep learning network,and greatly improves the accuracy of the interference detection when realizing intelligent and automatic and reducing artificial error.Finally,for the interference detection problem in the air-ground intercom system of the civil aviation,this paper puts forward an interference detection technology by the existence of interference detection and interference modulation recognition,and then simulates the interference detection method and modulation recognition method for various interference in the air-ground intercom system of the civil aviation.The result of the simulation analysis shows that the technology has high accuracy of the interference detection for various interference in the air-ground intercom system of the civil aviation,and has good recognition accuracy under low SNR to recognize the various complex signals.