An Accelerator to Additive Homomorphism to Handle Encrypted Data

Abstract

Homomorphic encryption provides a way to operate on the encrypted data so that the users can be given with the maximum feasible privacy. Homomorphic encryption is a special kind of encryption mechanism that can resolve security and privacy issues with rich text. Research gap is the performance overhead associated with this which poses a barrier to the real time implementation of this scheme. The objective of this work is to implement an algorithm to achieve increased performance and faster execution when compared with a classical cryptographical algorithm, the Paillier Cryptographical Algorithm, which is predominantly used to achieve additive homomorphism and analyse the performance gain obtained by this algorithm. The same algorithm is also integrated into an encrypted database application, CryptDB, developed by the MIT, as a replacement to the Paillier algorithm used in the application. The derived algorithms are 2600 time faster in key generation, 5 lakh times faster in encryption, and 3500 times faster in decryption, when compared with the Paillier algorithm.