Continuously Non-malleable Codes from Authenticated Encryptions in 2-Split-State Model

Abstract

Tampering attack is the act of deliberately modifying the codeword to produce another codeword of a related message. The main application is to find out the original message from the codeword. Non-malleable codes are introduced to protect the message from such attack. Any tampering attack performed on the message encoded by non-malleable codes, guarantee that output is either completely unrelated or original message. It is useful mainly in the situation when privacy and integrity of the message is important rather than correctness. Unfortunately, standard version of non-malleable codes are used for one-time tampering attack. In literature, we show that it is possible to construct non-malleable codes from authenticated encryptions. But, such construction does not provide security when an adversary tampers the codeword more than once. Later, continuously non-malleable codes are constructed where an attacker can tamper the message for polynomial number of times. In this work, we propose a construction of continuously non-malleable code from authenticated encryption in 2-split-state model. Our construction provides security against polynomial number of tampering attacks and non-malleability property is preserved. The security of proposed continuously non-malleable code reduces to the security of underlying leakage resilient storage when tampering experiment triggers self-destruct.