Research On Efficient Parallel Searchable Encryption Schemes Against Inside Keyword Guessing Attack

With the rapid development and maturity of cloud computing and cloud storage,cloud-based data storage and usage is an inevitable trend.With the convenience of cloud computing and cloud storage,a series of data security problems came out.In the mode of ciphertext storage,that the user’s data are encrypted and uploaded to the cloud,the data security can be effectively protected.However,after storing encrypted data in the cloud,the server is faced with the problem of how to retrieve the encrypted data and ensure the security of user data.In this context,the searchable encryption technology came into being.The existing parallel searchable encryption schemes are difficult to take account of search efficiency and data security.To solve this problem,this paper studies parallel searchable encryption scheme considering both security and efficiency.Aiming at the problem that the single keyword searchable encryption schemes can not balance resisting inside keyword guessing attacks and efficient search,this paper proposes an improved parallel searchable encryption scheme,which encrypts keywords and encryption times together to generate keyword ciphertexts,builds the relationship between a keyword and its multiple ciphertexts,realizes parallel search and improves search efficiency.In the proposed scheme,the selection of the number of parallel search lines is considered.Through the keyword frequency vector,the server can obtain the number of matched keyword ciphertexts before searching,optimize the search efficiency and avoid the waste of additional computing resources.At the same time,according to the nature of the elliptic curve,the server’s multiple bilinear pairing operations are transformed into a single bilinear pairing operation in the process of data receiver generating the trapdoor,which further improves the search efficiency.In addition,a pair of public and private keys are added for the data sender and he adds its own private key when encrypting keywords with the public key of the receiver,so as to resist inside keyword guessing attacks.Theoretical analysis and simulation results show that the parallel search time of the proposed scheme tends to be constant and meets the semantic security of trapdoors.Moreover,it can resist inside keyword guessing attacks and has high data security.In order to expand the functions of multi-keyword search and keyword fuzzy search,a parallel searchable encryption scheme supporting multi-keyword fuzzy matching is proposed in this paper.In the proposed scheme,the keywords are first transformed into bigram vectors,then the bigram vectors are mapped into bloom filter by local sensitive hash functions to generate index vectors,and they are encrypted to realize fuzzy search.The receiver employs the same method to generate the trapdoor vectors and encrypts them to generate the trapdoors.We set a search line for each trapdoor,employ parallel search to realize multiple search lines and search the keyword ciphertext matching each trapdoor at the same time to realize multi-keyword search.The server preprocesses the keyword ciphertexts and employs the inverted index to store the corresponding relation table between the keyword ciphertexts and the document numbers,so as to improve the search efficiency and reduce the storage space of the ciphertexts.Simulation results show that the proposed scheme can not only improve the search efficiency,reduce the storage space of ciphertext,balance the search efficiency and computing resources,but also meet the semantic security of trapdoor and keyword ciphertext and can resist inside keyword guessing attacks.Moreover,it supports the transformation between complete matching and partial matching patterns of keywords to meet the specific needs of users.The parallel searchable encryption scheme proposed in this paper takes into account data security and data search efficiency,and has high application value in the fields of electronic medical records,blockchain and so on.