Research On Intelligent Signal Detection And Secure Transmission Based On MBM

Media-Based Modulation(MBM)technology is a new modulation technology which modulates the data information to the different channel state information(CSI)after leaving the transmitting antenna compared with traditional modulation technology.MBM technology has become an important issue because it makes full use of the characteristics of the wireless channel and has the advantages of low cost and high spectral efficiency.The receiver needs to obtain accurate CSI by channel estimation in traditional MBM signal detection which places high requirements on the channel estimation module.And due to the rapid development of the artificial intelligence,its potential in wireless communication is increasingly apparent,so this thesis first studies the intelligent detector for the MBM system.Then,the secure transmission of wireless communication has received great attention because information security is becoming more and more important,so this thesis also studies the secure transmission technology of the MBM system.Firstly,this thesis gives the basic principles of the required content.Including the basic model of the MBM system and several traditional signal detection algorithms,Support Vector Machine(SVM),K-Nearest Neighbor(KNN)and Deep Neural Network(DNN)which are often used for classification and the physical-layer secure transmission technology.Secondly,this thesis studies the intelligent detector for the MBM system.In order to apply the classification algorithm in machine learning to the MBM signal detection,this thesis studies the conversion process of the signal detection problem into the classification problem and obtains the signal detection classification model.Then this thesis studies three kinds of intelligent detectors based on the signal detection classification model: SVM detector,KNN detector,DNN detector.Among them,in order to be suitable for the MBM signal detection,this thesis studies the SVM detector for multi-classification,the KNN detector with lower computational time complexity,and the DNN detector with faster training speed.The performance of the above three detectors is verified by simulation.Finally,This thesis gives the expression of the security capacity of the system,obtains several methods to improve the security performance of the system,and gives the corresponding traditional secure transmission technology.Then a secure transmission technology based on dynamic mapping is proposed which is based on that both legal receivers and transmitters know the accurate CSI and the MBM system selects the mapping method according to the CSI channel quality information.This thesis proves the theoretical feasibility of the secure transmission technology based on dynamic mapping through theoretical derivation and simulation results and the practical feasibility by implementing this system based on field programmable logic gate array.