Evaluation of encrypted matching criteria using homomorphic encryption based FSM
摘要
This paper presents a novel approach to encrypt finite state machines (FSM) that process plaintext data using a homomorphic encryption layer. In the proposed model, the FSM, its internal states, and the output state are homomorphically encrypted while the input stream is clear plaintext. Thus, the proposed model presents a methodology to execute encrypted operations that represent confidential matching over plaintext data and produce encrypted results, with the main target of protecting the matching criteria confidentiality. The modular design feature of the proposed system allows seamless integration with other modules to process complex matching criteria. The analysis and experimental results show better systematic features, faster rules encryption and results decryption, smaller encrypted rule size, better communication bandwidth, and higher security strength compared to the state-of-the-art.