Research On Encryption Technology Based On Nonlinear Adaptive Decomposition

Information security plays a momentous role in various fields,especially trade secrets and military secrets.To be precise,the main purpose of cryptography is to deliver ciphertext instead of plaintext to users in an unsecured channel.The research of encryption technology is one of the means to realize the integrity and security of data transmission,which is to ensure the confidentiality and privacy of communication.In recent years,with the increasing essentiality of information security,computer cryptography has developed greatly,but there is still insufficient research on the encryption of nonlinear and non-stationary signals.Nonlinear,non-stationary signals are well processed due to nonlinear adaptive decomposition techniques such as empirical mode decomposition(EMD)and singular spectrum analysis(SSA).Therefore,the main research in this paper is to apply nonlinear adaptive decomposition techniques to cryptography.This dissertation will focus on the combination of traditional cryptography and nonlinear adaptive decomposition techniques.The main contributions and innovations are as follows:1.Although the encryption technology based on filter bank fully considers the time-frequency characteristics of the signal in encryption,the algorithm has much room for improvement in signal reconstruction accuracy.Therefore,this paper proposes a new encryption algorithm based on empirical mode decomposition.Empirical mode decomposition is an adaptive,nonlinear and non-stationary signal processing tool.The encryption algorithm based on empirical mode decomposition can be divided into two stages: diffusion and confusion.In the confusion phase,the intrinsic mode functions are grouped by correlation between the intrinsic mode functions.The structure,proposed by Min-Sung Koh,processes the grouped intrinsic mode functions and generated three subband signals which performs a specific downsampling operation.In the diffusion phase,chaotic modulation is performed on the subband signals generated in the confusion phase.The frequency response of the eigenmode function behaves like the output of a binary filter bank.Therefore,using empirical mode decomposition can make full use of the time-frequency characteristics of the signal.The experimental comparison shows that the structure of the encryption technology based on empirical mode decomposition achieves higher reconstruction accuracy.2.The intrinsic mode functions are generated by repeatedly interpolating the extreme points of the signal to generate the upper and lower envelopes,subtracting the envelope from the signal,and iterating until the termination condition is satisfied or the residual is monotonous,requiring more computation time.Since singular spectrum analysis is also a data-driven signal analysis tool,and Min-Sung Koh's downsampling structure is versatile,it can perfectly reconstruct and allow arbitrary tree structure.Therefore,the singular spectrum analysis is used to replace the empirical mode decomposition.An encryption algorithm based on singular spectrum analysis is proposed.This method achieves the same level of security performance and reconstruction accuracy of the encryption technology based on empirical mode decomposition,but the computational efficiency is greatly improved.