Public key encryption resilient to leakage and tampering attacks

In this work, we investigate how to protect public key encryption from both key-leakage attacks and tampering attacks. First, we formalize the notions of chosen ciphertext (CCA) security against key-leakage and tampering attacks. To this goal, we then introduce the concept of key-homomorphic hash proof systems and present a generic construction of public key encryption based on this new primitive. Our construction, compared with previous works, realizes leakage-resilience and tampering-resilience simultaneously but completely independently, so it can tolerate a larger amount of bounded-memory leakage and be instantiated with more flexibility. Moreover, it allows for an unbounded number of affine-tampering queries, even after the challenge phase. With slight adaptations, our construction also achieves CCA security against subexponentially hard auxiliary-input leakage attacks and a polynomial of affine-tampering attacks. Thus, to the best of our knowledge, we get the first public key encryption scheme secure against both auxiliary-input leakage attacks and tampering attacks.