Physical Layer Encryption Methods And Technologies Basing On Channel State Information Nonlinear Transformation

The remarkable advancement in communication technology has greatly changed the way people live their lives and contributes to a profound impact on social progress.Since encryption methods are the most effective way to achieve data security and protect privacy of legitimate user,it is significant to study on security communication technologies.Although network layer encryption methods basing on cryptography can guarantee information confidentiality,illegal users may intercept the signals through open wireless channel and crack information using increasingly powerful backend system.By contrast,physical layer encryption method could provide more reliable secrecy performance.However,the existing communication security technologies are affected by many factors,such as synchronization performance,sensitivity of initial value,power loss caused by encryption processing and secury communication distance.For this reason,this paper propose the physical layer encryption methods based on channel state information nonlinear transformation.By nonlinear transformation,channel transmission characteristic is changed so that the signal reception quality of illegal users is effectively deteriotated and meanwhile the difficulty of decoding the intercepted signals is increased.On the other hand,illegal user's decryption strategies for the above encryption methods are studied.A blind nonlinear identification and compensation method based on optimal group delay searching has been presented in this paper.The proposed encryption system's security performance is evaluated by comparison of the demodulation effect before and after nonlinearity compensation.The main researches of this thesis can be concluded as follows:(1)For the physical layer security issue,the basic ideas of carrying out nonlinear transformation on channel state information to realize physical layer encryption are discussed: the non-ideal transmission characteristics of communication system will bring in signal distortions.Their generation mechanism,models and identification methods are discussed.After theoretical analysis and simulation,the encryption ideas of implement nonlinear transformation on channel's group delay function and transmission function are established.Besides that,illegal user decryption strategies for the proposed encryption methods above are also discussed and a blind nonlinear identification and compensation method basing on best group delay searching is designed so that it can deal with the joint effect of nonlinear distortion and group delay distortion of encryption system.(2)For the physical layer security issue,a group delay function nonlinear transformation based encryption method(GDNT)is proposed.Its core principle is taking advantage of group delay function to generate nonlinear transformation pair so that the correlation of signals before and after encryption can reach its minimum value to meet the perfect secrecy(PS)condition.Suffered from inter-symbol interference,encrypted signal's constellation will be scattered and neither can become correct information through sampling judgement and coding.According to PS condition,the only way for illegal user to obtain original information is randomly guessing.GDNT makes use of signal correlation in frequency domain to form objective and derives nonlinear equations of encryption parameters in group delay function when objective reaches its minimum.Then group delay function is expressed by improved Fourier series model and iterative formula of its encryption parameters is given basing on weighted nonlinear least square algorithm(WNLS).Finally,the results of encryption and illegal user decryption experiments show that GDNT has a satisfactory encryption effect and reliable security.(3)For the physical layer security issue,a transmission function nonlinear transformation based encryption method is proposed.Its core principle is taking advantage of memory nonlinear transform pair(MNT)to change channel's magnitude-phase transmission characteristics so that nonlinear spectral regrowth is generated within original signal band.These nonlinear distortions will reduce original signal to interference ratio and deteriorate illegal user's signal quality.MNT creates nonlinear transformation pair with power series model of difference frequency component of third order intermodulation.Then optimizing strategy for control band of encrypted signal is provided.In addition,memory effect is brought into MNT basing on phase filtering,where group delay is represented by oscillating wave function and pseudo random sequence respectively.The design steps of phase encryption filter in frequency domain are given.At last,the encryption effect and legal user's loss of signal to noise ratio for MNT encryption system are shown to evaluate its performance.(4)For the sake of study illegal user decryption strategies for MNT encryption system,this paper analyzes its joint effect of group delay distortion and nonlinear distortion and provides a best group delay searching based nonlinear system blind identification and compensation method(GSIC)as illegal user decryption method.GSIC takes the instantaneous power residual error between weak signals and output signals when strong signals going through nonlinear behavior model as objective function.The linear equations of kernels of nonlinear behavior model are derived when objective reaches its minimum value.After that,weighted iterative improvement algorithm(WII)is used to solve the ill-conditioned equations.Through verification experiments of compensating nonlinearities of simulation signals and signals from real world front-end,GSIC's nonlinear distortion suppression performance is satisfactory.Finally,GSIC is used as an illegal user decryption method for MNT encryption system to evaluate GSIC's decryption performance and at the same time,the necessities of MNT system spectrum optimization strategy and deep memory effect are verified.