Research On Image Encryption And Secure Transmission Based On Tensor Product Compressive Sensing

The popularity of the Internet of Things and the development of 5G technology promote people's dependence on the network.More and more data are transmitted over the network.As a main information carrier,digital image occupies the core part of data transmitted in the network,and its security has been widely concerned.Traditional encryption technologies for text data cannot effectively encrypt images.Therefore,a series of image-based encryption technologies are proposed.However,with the increasing number of images which are calculated,stored and transmitted on the network,the pressure of the network is also increasing.The high-efficiency image encryption technologies have become the top priority of people's researches.The compressive sensing technology realizes sampling,compression and encryption simultaneously.Once proposed,it has been widely used.Whereafter,in view of the storage consumption of large measurement matrix in compressive sensing technology,researchers put forward the semi-tensor product compressive sensing technology and P-tensor product compressive sensing technology.These two technologies realize the purpose of processing larger images with smaller measurement matrices and greatly improve the energy consumption of these systems.The main work of this thesis includes the following three aspects:(1)By deeply understanding the existing image encryption schemes based on compressive sensing technology and P-tensor product compressive sensing technology,summarizing the advantages of these schemes,and analyzing the defects of these schemes,we find out the aspects that can be further optimized and improved,and lay a groundwork for the future work.(2)A multi-image encryption scheme based on P-tensor product compressive sensing is proposed.The scheme takes full advantage of the P-tensor product operation in matrix multiplication,and uses the same measurement matrix to sample images of different sizes.Then a secure and low energy consumption matrix updating method is designed to update the measurement matrix.The scheme not only ensures the image security,but also greatly reduces the resource consumption of both communication parties.The parallel processing of multiple images enhances the timeliness of the scheme.(3)A secure image transmission scheme based on P-tensor product compressive sensing is proposed.The scheme takes into account the requirements of confidentiality,integrity and authenticity of the images transmission in the insecure channels,and ensures the secure transmission of the images through three modules.Firstly,P-tensor product compressive sensing technology and confusion,diffusion operation are used to encrypt the images to ensure confidentiality.Then authentication is carried out on the measurements generated in the previous step to ensure the integrity of images.Finally,under the inspiration of asymmetric key encryption technology and public key authentication technology,a secure key transmission scheme is designed.While the two parties share the keys,the identity of both parties can be verified and ensure the authenticity of the transmitted information.