Public-key encryption for protecting data in cloud system with intelligent agents against side-channel attacks

Public-key encryption can be used to protect the sensitive data in cloud system with intelligent mobile agents facilitating better services. However, many public-key encryption schemes do not resist the side-channel attacks which can be applied to the encryption instances implemented on a chip or cloud to obtain partial information leakage about the secret states, as the traditional security model of public-key encryption does not capture this kind of attacks. Also, the adversary can inject fault to tamper with the secret key and observe the output of the public-key encryption scheme under this modified key which is called "related-key attacks". Inspired by these, the models of key-leakage attacks and related-key attacks are formalized, respectively. In this paper, we present a method to construct public-key encryption schemes against both weak key-leakage attacks and linear related-key attacks from extractable hash proof systems (EHPS or XHPS). Specifically, we first transform ABOEHPS to weak leakage-resilient ABOEHPS and add Key Homomorphism and Fingerprinting properties to it. Then, based on this new ABOEHPS, we construct weak leakage-resilient adaptive trapdoor relation with these two properties and public-key encryption schemes against both weak key-leakage attacks and linear related-key attacks can be constructed from it. Moreover, we propose a public-key encryption scheme against both adaptive key-leakage attacks and linear related-key attacks.