| The rapid progress of information technology makes it possible to generate big data. After the cloud computing and Internet of things big data becomes another hot words, which is obtained great popularity and attention from all walks of life. Big data has the so-called characteristics of "4V". Volume presents massive scale of data. Variety is various types of data. Velocity means fast transfer of data and dynamic system of data. Value expresses huge value of data. The most core value of big data lies in storing and analyzing the massive data. Among them the most important problem is the security of it. The traditional data encryption technology is difficult to satisfy the requirement in the speed of encryption, so we need to study the encryption technology for the big data.This paper gives three data encryption algorithms for the big data and compare these algorithms according to the characteristics of big data. Because of the feature of big data which needs quickly and efficiently encryption, we improve the algorithm, in order to increase the encryption speed of these algorithms. The main research content and innovation points are as follows.On the basis of analyzing present existing data security challenges and research progress, three algorithms for encrypting the big data are given. Also these three algorithms for their shortcomings are improved to get modified linear mixed encryption algorithm and modified hybrid key encryption algorithm.Improving elliptic curve encryption algorithm includes three aspects. First, improve the bottom operations. Give the speed of point adding and doubling operation in three kinds of coordinate systems, namely projective coordinate system, Jacobi coordinate system and hybrid coordinate system of these two coordinates. In projective coordinate system executing point adding operation and in Jacobi coordinate system executing doubling operation can improve the speed of these two operations. Second, improve high level operations. This paper gives the speed of doubling operation of the binary square multiplication algorithm, window algorithm and NAF algorithm. The speed of the NAF algorithm is fast. Third, improve the finite field. In this paper we compare the computing speed of ECC in the finite field of p>3 and p=2. It concludes that the speed of point adding and doubling operation in the finite field of p=2 is slightly faster than p>3. It can improve the computing speed of elliptic curve encryption algorithm.On the study basis of the first two parts, the big data’s encryption algorithm is improved. By using the improved ECC and big data’s encryption algorithm, on the platform of Shuguang’s XData-Hadoop big data software we realize it and analyze the algorithm. The analysis results show that the improved big data’s encryption algorithm can resist exhaustive key search in some extent. Its security is better and the speed of encryption is fast. |
