A novel implementation of parallel homomorphic encryption for secure data storage in cloud

Abstract

Due to the digitization of data in large scale and varying requirements of the organizations to serve different stakeholders, there is a need to compute, store and analyze large amount of data often in remote servers efficiently. However, each individual entity is not equipped with the right amount of resources to perform necessary computations on large data. Cloud computing provides a platform to compute, store and share data remotely for heterogeneous entities and provides services to them according to their demand. One of the major challenges in cloud computing platform is to ensure security and privacy of the data. The major limitation of applying existing cryptosystem such as AES, 3-DES, etc. is the need of decryption of ciphertext for outsourced computations that may violate data security. Homomorphic encryption is a solution which allows performing computations directly on cipher text without decrypting of the ciphertext. In this paper, we present an efficient model to implement different algorithms related to benchmark computations on ciphertext derived using the Dijk, Gentry, Halevi and Vaikuntanathan's homomorphic encryption scheme (DGHV scheme-2010). We have also executed 'ciphertext refresh' procedure at Keg Generation Server (KGS). In addition, we present parallel implementation of these algorithms and provide a comparative performance study. � 2017 IEEE.