Research On Pilot Spoofing Attack Detection And It's Protection Technology

The surging data traffic and scarce spectrum resources have brought great challenges to 5G mobile communication system.Technologies such as massive MIMO and beamforming can significantly improve the spectral efficiency and capacity,on the premise that accurate channel state information(CSI)can be acquired.The general method for getting CSI is the pilot-based channel estimation,which however may be easily spoofed or interfered by an eavesdropper,resulting in unexcepted channel estimation errors.The CSI errors not only damage the communication performance,but also cause security problems like information leakage.The existing detection methods for pilot spoofing attacks have some disadvanges like pilot structure redesign,high training complexity,or requiring prior information.This thesis tries to enhance the detection performance and mitigate the attack effects by taking physical layer authentication and security protection technologies.The main contributions are as follows:In order to detect the pilot spoofing attack in the time division duplexing(TDD)multi-antenna system,a multi-node assisted physical layer detection method is proposed.Firstly,the legitimate user is positioned by using the RSS(Reiceived Signal Strength)measured by auxiliary nodes,then the attack detection problem is transformed into the user's position authentication within the hypothesis testing model.To improve the positioning and detection accuracy,the deployment of auxiliary nodes is optimized.The expressions of detection probability is derived based on probability theory.Simulation shows that compared with the traditional Energy Ratio Detector(ERD)and Two-Way Training Detector(TWTD),the proposed method can improve the detection probability by 58%or 25%,without destroying the pilot structure or increasing the training complexity.However,complex channel environment may cause performance degradation.In order to improve the performance of pilot spoofing attack detection in complex channel environment,a physical layer detection method based on improved orthogonal pilot training is proposed.Legal users use different powers to insert orthogonal pilot randomly into conventional pilots,which will increase the differences with the eavesdropper's pilot sequences,thereby improving the base station's detection capability.simulation shows that compared with the traditional orthogonal pilot training detection method,the detection probability of the proposed method is significantly improved,which can still reach 98%even if the eavesdropper fully masters the pilot training rules.In order to mitigate information leakage caused by pilot spoofing attack,a joint optimization scheme for orthogonal pilot detection and physical layer security transmission is proposed.Under different detection results,the base station will design different secure transmission schmes based on maximum ratio transmission(MRT)or zero-forcing(ZF)beamforming.In this scheme,the detection performance is included in the transmission performance evaluation index.In order to further meet the quality of service(QoS)and security requirements of the communication system,the optimized detection threshold and the length of pilot are derived.Simulation shows that the above parameters directly affects the detection performance,and further affects the transmission performance.By adopting the optimized parameters,the system can reach the maximum weighted effective rate,obtains better transmission quality.