Research On Two Hybrid Encryption Algorithms For Big Data

In recent years,with the booming development of the Internet,people are more dependent on the network for communication,shopping and other behaviors.This leads to an exponential increase in data volume,hence the emergence of big data.Big data has the characteristics of openness,interactivity and globality,and has great value.Some people even compare big data to oil and gold.Big data has become the focus of companies,enterprises and countries.But big data brings new challenges as well as opportunities.At present,big data-oriented encryption algorithms have disadvantages such as short key length and slow encryption speed.How to protect data security in big data environment has become one of the most important problems.Only by solving the bottleneck of information security can information technology develop rapidly and widely.After studying the existing data encryption algorithms and literatures,this thesis innovatively proposed the improved DES algorithm based on stream cryptography and the hybrid encryption algorithm based on hyperchaos and conic curve.Both algorithms have improved their security.The two encryption algorithms are computed in parallel on Spark platform to improve the encryption speed.The specific research contents of the thesis are as follows:(1)an improved DES encryption algorithm based on stream cipher is proposed.Through the research on the key generation process in DES encryption algorithm,it is found that there is a strong correlation between DES sub-keys,Therefore,a DES improvement scheme based on stream cipher is proposed.The improved scheme generates pseudo-random sequence through RC4 algorithm,which is used as the sub-key of DES encryption to encrypt the plaintext of DES encryption algorithm.The randomly selected key using RSA encryption is transmitted with the ciphertext for decryption of the ciphertext.The improved scheme makes every plaintext group of DES use different encryption keys for encryption,at the same time the key length increased to 768 bits,so that the number of breaking a group of encryption keys to 2768,greatly increased the difficulty of DES decoding.(2)a hybrid encryption algorithm based on hyperchaos and conic curve is proposed.By analyz:ing the shortcomings of the encryption algorithm designed by one-dimensional chaotic system,such as short key length and small key space,a four-dimensional hyperchaotic encryption algorithm is proposed to increase the key space.First,two hyperchaotic systems are used to generate a hyperchaotic sequence without correlation.Then the plaintext and hyperchaotic sequences are encrypted for the first time by performing xor operations.Finally,the encrypted ciphertext is used as the plain text of conic curve encryption for secondary encryption.Through experimental comparison and analysis,on the one hand,the algorithm has the advantages of large key space,good ciphertext statistical characteristics and high key sensitivity.On the other hand,there is no direct connection between the plaintext and ciphertext after double encryption,and the key sequence cannot be cracked by selecting special plaintext and ciphertext pairs.Besides,the nonlinear operation in the algorithm can resist the selective plaintext attack and improve the security of the algorithm.(3)two encryption algorithms were tested in parallel through Spark.Because of the advantages of Spark in parallel computing,such as fast,simple and easy to use,universality and running mode.The improved DES encryption algorithm based on stream cipher and the hybrid encryption algorithm based on hyperchaos and conic curve are analyzed by parallel algorithm theory.The experimental results show that for the same size data set,the encryption time decreases with the increase of the number of nodes.In terms of the acceleration ratio,with the increase of the number of nodes,the acceleration ratio also increases gradually,achieving the purpose of improving the encryption speed.