Search Encryption Algorithm For Multiple Data Owners Against Keyword Guessing Attack

With the rapid development of cloud computing technology and the many advantages of the cloud systems like low cost,flexible,rapid retrieval etc.,more and more data users are beginning to use the cloud to store and share great amounts of data.This has the advantage that data owners can be relieved from the burden of huge data storage and maintenance so as to enjoy on-demand highquality data storage services.However,the server may be vulnerable and the stored data may be leaked to a malicious attacker.So,data owners may encrypt data before outsourcing them to the cloud.This makes how to efficiently search over encrypted data an important problem.In this case,a new encryption primitive was proposed,which was called searchable encryption technology.It subverts the defect that the traditional public key encryption system cannot directly operate the ciphertext.However,the searchable encryption schemes proposed are only for single data owner currently,and these schemes are vulnerable to keyword attacks.Based on the above problems,this paper studies the content of multiple data owners search and against keyword guessing attack.The specific contributions are as follows:(1)Locally sensitive hashing algorithms and bloom filters are introduced into the field of keyword search: in order to implement fuzzy multi-keyword search,keywords can be divided into binary groups by binary coding and binary vectors can be further generated.According to the characteristics of the binary system vector,a misspelled keyword can still be represented by a binary vector,and the misspelled keyword can be highly tolerated.At the same time,in order to greatly improve the efficiency of data search,this solution can complete the search of multiple data owners by only one retrieval.(2)A new encryption algorithm is proposed,and referred to as MDO-PEKS: a new protocol for generating random keys is proposed,which can support different data owners to have different keys.In this way,the data security of each data owner is protected.In addition,it is difficult to manage the ciphertext format of the plaintext after it is encrypted with different keys.But MDO-PEKS solves this problem and enables the ciphertext to be managed uniformly,which also reduces the management cost and improves the search efficiency.In addition,MDO-PEKS can solve the bilinear Diffe-Hellman problem.(3)A scheme that can against keyword guessing attack is proposed,and referred to as AKGA-PEKS: based on MDO-PEKS,the index and trapdoor are obscured,and a new multiple data owners framework supporting fuzzy multi-keyword search is proposed.MDO-PEKS not only againsts keyword guessing attack,but also retains the powerful search capabilities of fuzzy,multi-keyword,multiple data owners.It has been proven that this solution can also solve the bilinear Diffe-Hellman problem,further ensuring the security of the data.