Efficient and Privacy-Preserving Outsourced Calculation with Multiple Keys Using Chinese Remainder Theorem

We propose a new scheme for privacy-preserving outsourced calculation. With the help of this scheme, a user on a local host can securely transfer his sensitive information to a cloud platform for processing and storage avoiding disclosing the original information to the unreliable cloud and prying eyes. In addition, the Cloud Platform (CP) can carry out specifical operations on ciphertexts with the help of a Cryptographic Service Provider (CSP) according to the user's requests and send the results to user in the form of ciphertext. Furthermore, data encrypted with multiple keys can also be calculated together, and the computed results will be decrypted by an authorized receiver. We not only implement addition and multiplication operations on ciphertexts, but also design an efficient comparison protocol on ciphertexts (the operations are all in encrypted domain). We then show the security of our encryption and the suggested secure computation sub-protocols. We use a one-way trapdoor permutation and a symmetric somewhat homomorphic encryption as the framework's building blocks rather than the fully homomorphic encryption (FHE) technique. Finally, we use simulations to show the effectiveness and extraordinary efficiency of our suggested scheme.