An Oracle Scheme for Multi-Source Data Privacy Protection and Source Authentication

With the continuous development of blockchain technology, massive off-chain data is mapped on the blockchain, ensuring the authenticity and privacy of on-chain data and off-chain data is a significant challenge. To solve this question, many studies use oracle to provide secure and reliable data for blockchain applications. Existing oracle schemes can protect the privacy of single-source data and prove the authenticity of private data sources to the third party. However, when handling multi-source data, these schemes require multiple executions to process and verify all data. We propose an optimized oracle scheme based on the “TLS-MPC” framework to improve efficiency. Firstly, we optimize the handshake process by dividing the $\mathrm{n}$ servers into $\mathrm{t}$ clusters and use the session ticket to reduce the number of MPC executions during the three-party handshake. As a result, most servers within each cluster run a fast three-party handshake by session ticket. Secondly, the prover runs two-party computation with the verifier to generate the queries and sends them to each data source to get the multi-source data. Then we design a constructable zero-knowledge proof system. Prover will inputs the multi-source data into the system to generate the proof value with a joint computation circuit. Finally, verifier will check the result sent from prover after the zero-knowledge proof is completed. Comparing with the DECO, our scheme is more efficient.