Design And Implementation Of Distributed Image Encryption Algorithm Based On MapReduce

With the advent of the big data era,the image information on the Internet has rapidly increased,the presentation of big or even large-scale image has posed a new challenge to image encryption technology.The biggest difficulty of large-scale image encryption is the efficient decrease method of encryption processing time.The existing image encryption technologies usually use heuristic algorithm or reduce the complexity of algorithm to reduce the time cost.It is inevitable that using heuristic algorithm may leads to local optimum,while reducing the complexity would reduce the secure of the image encryption.So now,for all the images encryption algorithm,if we want to maintain the image encryption effect under the premise of such a large-scale image data,the computation time is huge.So it is an imperative requirement in this field that we need to pose a massive image data encryption algorithm which firstly it can keep the effect of the encryption algorithm and the encrypted security,at the same time it also can reduce the time cost when the encryption speed improved.Aiming at these problems above,we consider from parallelization of large-scale Image Encryption algorithm,on the premise of effectively improve the real-time performance of the algorithm and guarantee the security of Image Encryption,then I design and realize a real-time Distributed Image Encryption algorithm based on MapReduce---DIE-MCS(Distributed Image Encryption based on MapReduce and Chaotic Systems)algorithm.In order to effectively improve the computing speed of large-scale image encryption algorithm,this paper draws on Google MapReduce parallel processing ideas,it is based on the Linux MPI cluster environment by using MapReduce-MPI parallel library,then divide the large-scale image into several computing nodes to deal with a part of sub-images.At the same time,this paper parallel encryption processing which greatly has reduced the encryption processing time to ensure the real-time property of the algorithm;In order to meet the real-time requirements,it also ensure the safety of the image encryption algorithm.In order to effectively ensure the security of the encryption algorithm,we adopt the most stable image encryption algorithm of the chaotic system in the current image encryption algorithm,which effectively enhances the sensitivity of the DIE-MCS algorithm.Based on the experimental results of DIE-MCS algorithm in terms of security,real-time and sensitivity etc.,we get the following conclusions:DIE-MCS Algorithm use MapReduce-MPI parallel computing framework.The natural advantage of distributed processing is used to accelerate the encryption algorithm effectively.Experiments are performed in a cluster of 16 cores,and the data encryption speedup ratio can reach 15,and as the number of computing cores Increasing,the speedup ratio also increases,it is suitable for the more demanded real-time applications field.The encryption part of the DIE-MCS algorithm uses a chaotic system to encrypt the image.Since the chaotic system itself can provide a large key set,the encrypted data cannot be cracked in the opposite direction.At the same time,the traversal characteristics of chaos encryption algorithm itself can make the key distribution more random and uniform,further improved the security of DIE-MCS algorithm.DIE-MCS algorithm utilizes the high sensitivity of chaotic system to initial conditions,and the NPCR and UACI indices reach 99.93%and 33.23%respectively,which can effectively resist differential attacks.Based on the comprehensive Linux MPI cluster environment to build distributed encryption computation,DIE-MCS algorithm has a nice portability and expandability and can be efficient deployed rapidly and applied comprehensively.In conclusion,the DIE-MCS algorithm proposed in this paper has achieved a good balance in terms of security,real-time and sensitivity and it has good application prospects.At the same time,the research work in this paper has certain theoretical value,which has referential value for similar work.