Research And Implementation Of The Hybrid Encryption Algorithm In The Cloud Computing

In recent years, with the rapid improvement on the internet of things, mobilecomputing and big data stored, much more large-scale IT companies and researchinstitutions have done more deeply research into cloud computing. Cloud computingis known as super-speed computing power, super-large storage capacity, and suppliesdue to specific requirements, so that it is regarded as the framework of the coretechnology in next generation computer network. When mass data are put into thecloud computing system, more focuses will be set on the security. As the storeservice provided by cloud computing is totally unseen, it is difficult for commonusers to know the detail information of the data they stored, so that the non-sense oftrust has blocked the improvement of cloud computing, therefore, studying the datasecurity based on cloud computing is of vital importance.This thesis starts with the cloud computing infrastructure, named as Hadoop,which is developed by Apache foundation, and then, we introduce the HDFSdistributed file system, as well as the MapReduce parallel computing framework inthe Hadoop. The HDFS referred to has the advantage of high ability offault-tolerance, and strong expansibility, and it’s able to lay out onto the cheaphardware. Meanwhile, it also offers some redundant data to guarantee the stored datastable. On the other hand, MapReduce computing framework could shield the detailsin bottom platform, resulting in confidence for developers to concentrate on the codecompiling and designing.For the data security, this paper mainly discusses the structure improvement ofencryption algorithm. Taking algorithms of DES，AES and RSA for example, thispaper firstly analyzes traditional symmetric encryption algorithm and asymmetriccryptographic algorithm and compares the advantages and disadvantages, thendescribes whether the existed hybrid encryption algorithm is fit for the cloudcomputing model, in addition，concludes the algorithm structure of linear ADEShybrid encryption algorithm, double ADES hybrid encryption algorithm, doubleADES hybrid encryption algorithm with interference, and multi-hybrid encryptionalgorithm with interference. In addition, this thesis explains the similarities anddifferences theoretically among the four algorithms involving above in ways of keylength, plaintext sensitivity, key sensitivity, attacking and so on. After that, this paper adds the four hybrid encryption algorithms into Hadoopby means of MapReduce computing framework, and divides the data into elementaryunit for processing, say64MB,32MB,16MB,8MB,4MB, and2MB, then conductsencryption and decryption experiments with different data unit.During the experiments，the following records are noted：encryption runningtime，deciphering time，shortest time for single data block running and the longesttime for single block running. This paper analyzes the running time of each hybridencryption algorithm under different data blocks firstly, and then compares the leastrunning time of each algorithm, with the addition of the theoretical analysis,concludes the compatibility of different encryption algorithm in cloud computingenvironment at last.