SEDRES: SECURED DATA RETRIEVAL SERVICE

Abstract

Cloud service model can provide flexible storage place to store electronic data cost-effectively. However, storage of sensitive data in the third party cloud domain may violate the privacy of data. For this reason, encryption logic needs to be applied before uploading it to cloud storage. Use of encryption logic may violate the availability of data. It is now becoming a big concern to retrieve data securely from the cloud storage. This paper considers a remote application environment, where cloud data retrieval service must be secured to preserve the privacy of sensitive data. Here multiple items are searched in a single query to make the procedure faster, and the only specific number of relevant records can be retrieved at a time to save bandwidth cost. Encryption logic is proposed in such a way that encrypted cloud data could be processed without decryption. A secured retrieval service is proposed here to ensure data availability at any time. Hybrid encryption logic is designed based on AES cipher and homomorphism properties. Limitations of fully Homomorphic encryption are avoided in proposed logic. Analytical study of AES cipher shows its efficiency to encrypt electronic data and reply vectors. Theoretical analysis of homomorphism shows its efficiency to secure searchable index, items, and query vectors required in cloud data retrieval service.