PAESS: Public-Key Authentication Encryption With Similar Data Search for Pay-Per-Query

In recent years, many cloud service providers adopt the pay-per-query model to offer paid search services to the public. The data owner rents the resources of cloud service providers and charges the data user a fee based on the data volume to be queried. While this commercial model offers flexibility, convenience, and cost-effectiveness, it comes with a significant vulnerability to data breaches. Public-key authentication encryption with keyword search (PAEKS) is a technology which is well applied in the pay-per-query model. But there is no PAEKS scheme applicable to this scenario. For this purpose, we present public-key authentication encryption with similar data search for pay-per-query (PAESS) and construct the first idiographic scheme PAESS-I. PAESS-I utilizes Shamir secret sharing and locality sensitive hashing to implement similar data search, has the verifiability of results, adds the charge function to prevent cloud servers and data users from colluding to deny deductions. We propose the second scheme PAESS-II based on PAESS-I, which is a mobile-friendly lightweight PAESS. Our second scheme operates in the pay-per-query model without pairing and exponential operations. PAESS-I satisfies ciphertext indistinguishability and trapdoor indistinguishability, and sacrifices the computational performance in favor of the pay-per-query model. The optimized PAESS-II is resistant to adaptively-chosen-targets attack, and satisfies ciphertext indistinguishability and trapdoor indistinguishability. PAESS-II distinguishes itself from other existing similar schemes by having the same characteristics as PAESS-I, along with the benefits when it comes to the calculation cost.