Research On Physical Layer Security Technology For Pilot Spoofing Attack

Due to the openness of wireless communications,information is vulnerable to eavesdropping or malicious attacks by illegal users.In current wireless communication systems,encryption is commonly used to protect information security.However,with the improvement of computing technology and the development of decryption strategy,any encryption algorithm based on computational complexity can be brute forced theoretically,resulting in serious security risks in the transmission of signals.By exploiting the inherent random characteristics of wireless channels,the physical layer security technology could achieve the secure transmission of wireless signals.Such technology neither relies on computational complexity of encryption algorithms nor requires the distribution and management of keys.Therefore,it has attracted widespread academic attention.Obtaining accurate Channel State Information(CSI)is vital for physical layer security.CSI is usually estimated by employing pilot training.However,since the codebook of pilot training sequences is generally public,it can be easily obtained by eavesdroppers.If the eavesdroppers send the same pilot signals to the legitimate users,the estimation of CSI would be changed,resulting in information leakage in the following transmission,i.e.,under the Pilot Spoofing Attack(PSA).PSA can seriously threaten the physical layer security in wireless communications,so we study the PSA detection problem and the subsequent transmission optimization methods.The contributions of this thesis are summarized as follows:1)Firstly,to address the problem that PSA exists in Multiple-Input Single-Output(MISO)systems,this thesis presents a novel Three-Phase Uplink Training(TPUT)scheme which could precisely estimate the CSI of the main channel and the eavesdropping channel.Since the existence of PSA will change the statistic features of the eavesdropping channel CSI,this thesis constructs a binary detection problem to achieve PSA detection.To further consider that noise variances cannot be pre-determined in practice,a Generalized Likelihood Ratio Test(GLRT)detector is also proposed.Simulation results show that the proposed detector can achieve reliable PSA detection when noise variances are unknown.2)Secondly,this thesis proposes a novel "Two-Step Uplink Training(TSUT)" scheme for Intelligent Reflective Surface(IRS)-MISO systems with PSA.The scheme can accurately estimate the CSI of the backscattering main channel and the backscattering eavesdropping channel even when the reciprocity of channels is not satisfied.This thesis also analyzes the effects of user's transmitted power and the prior information of the channel and noise on the performance of PSA detection,thus a Rao detector along with a Locally Most Powerful(LMP)detector are proposed.Simulation results show that the performance of the Rao detector and the LMP detector are independent of the prior information of the channel and noise.The two detectors can accurately detect PSA when the eavesdropping signals are provided with high and low transmitting powers,respectively.3)Thirdly,for the scenario where the MISO system has a single-antenna eavesdropper,this thesis analyzes the effect of channel training time on the system transmission performance in the artificial noise transmission scheme.Since there is a tradeoff between channel training time and information transmission time under the total time constraint,this thesis derivates a closed-form expression of the traversal secret rate,and designs the optimal time and power allocation strategies by regarding the closed-form expression as an objective function.4)Finally,considering that there is a single-antenna eavesdropper in the IRS-MISO system,this thesis proposes a joint design scheme of transmitter beamforming and IRS reflection coefficients to achieve safe information transmissions on the physical layer.Furthermore,this thesis optimizes the system transmission efficiency under the time constraint,and derives closed-form expressions of the upper bounds of the secret rates for Monostatic Backscatter Channel(MBC)and Bistatic Backscatter Channel(BBC).After that,an optimal time allocation strategy is designed by regarding the closed-form expression as an objective function.