Analysis of Lattice Reduction Attack against the Somewhat Homomorphic Encryption Based on Ideal Lattices

Author(s):  
Masaya Yasuda ◽  
Jun Yajima ◽  
Takeshi Shimoyama ◽  
Jun Kogure
2014 ◽  
Vol 8 (3) ◽  
Author(s):  
Masaya Yasuda ◽  
Kazuhiro Yokoyama ◽  
Takeshi Shimoyama ◽  
Jun Kogure ◽  
Takeshi Koshiba

AbstractIn this paper, we revisit the fully homomorphic encryption (FHE) scheme implemented by Gentry and Halevi, which is just an instantiation of Gentry's original scheme based on ideal lattices. Their FHE scheme starts from a somewhat homomorphic encryption (SHE) scheme, and its decryption range is deeply related with the FHE construction. Gentry and Halevi gave an experimental evaluation of the decryption range, but theoretical evaluations have not been given so far. Moreover, we give a theoretical upper bound, and reconsider suitable parameters for theoretically obtaining an FHE scheme. In particular, while Gentry and Halevi use the Euclidean norm evaluation in the noise management of ciphertexts, our theoretical bound enables us to use the ∞-norm evaluation, and hence it helps to lower the difficulty of controlling the noise density of ciphertexts.


Author(s):  
Desam Vamsi ◽  
Pradeep Reddy

Security is the primary issue nowadays because cybercrimes are increasing. The organizations can store and maintain their data on their own, but it is not cost effective, so for convenience they are choosing cloud. Due to its popularity, the healthcare organizations are storing their sensitive data to cloud-based storage systems, that is, electronic health records (EHR). One of the most feasible methods for maintaining privacy is homomorphism encryption (HE). HE can combine different services without losing security or displaying sensitive data. HE is nothing but computations performed on encrypted data. According to the type of operations and limited number of operations performed on encrypted data, it is categorized into three types: partially homomorphic encryption (PHE), somewhat homomorphic encryption (SWHE), fully homomorphic encryption (FHE). HE method is very suitable for the EHR, which requires data privacy and security.


Sign in / Sign up

Export Citation Format

Share Document