Study And Implementation Of Order-Preserving Encryption In Cloud Computing Environment

With the rapid development and wide application of cloud computing,the privacy preserving of user data in untrusted cloud computing environment has gradually aroused people's attention.Currently,one of the major ways to protect these data is to utilize encryption algorithms.Data is encrypted first and then are stored into database in cloud.However,most of existing encryption schemes do not support direct operations over ciphertexts,which weaken the advantages of cloud computing.Therefore,it is a research hotspot and an urgent problem to be solved to support operation and retrieval of ciphertexts.Order-Preserving Encryption scheme can retain the order information of plaintext when data is encrypted,so users can obtain the order of ciphertext directly without decrypting.mOPE,proposed by Raluca Ada Popa in 2013,encodes ciphertext data depending on the structure of balanced binary search tree(AVL).This scheme supports arbitrary data type for encryption,but the operation of inserting or deleting is inefficient.Firstly,this paper proposes a generalized mutable Order-Preserving Encoding scheme,gmOPE,by improving the encoding adjustment mechanism of mOPE.gmOPE replaces the original balanced binary tree with the generalized balanced binary search tree(AVL-N)and introduces a new rebalance adjustment strategy.This new scheme ensures that ciphertext preserves the order information of the plaintext and it effectively reduces the frequency of rebalancing operations and improves the overall efficiency of Order-Preserving Encoding scheme.Then,to solve the problem of high frequency interaction between the client and server and the corresponding large amount of communication time,the Layered general mutable Order-Preserving Encoding(LmOPE)scheme is proposed,and the LmOPE scheme reduces the interactions between the client and server effectively,and further improves the performance.Finally,Homomorphic mutable Order-Preserving Encryption(H-mOPE)scheme is implemented by combining the Floating-point Fully Homomorphic Encryption(FFHE)with LmOPE.Not only can H-mOPE scheme directly compare the order of ciphertext but also execute operation of addition and multiplication on ciphertext for arbitrary times,and the introduction of homomorphic encryption algorithm does not have an obvious negative impact on performance of HmOPE.The theoretical analysis and experimental results show that gmOPE and LmOPE effectively reduce the extra expense which resulted from re-coding and interaction between the client and server,and efficiency is improved.In the whole,H-mOPE scheme,which developed by combining LmOPE with homomorphic encryption technology,not only maintains efficiency of the algorithm but also provides more operated methods for ciphertext.