Research On Privacy Protection Of Trusted Input Of Smart Contract Based On Zero Knowledge Proof

Smart contracts add programmability to blockchain.With smart contracts,blockchain users can accomplish more functions.After a smart contract is deployed to the blockchain,the user can request the service provided by the contract,and if the conditions set by the contract are met,the contract is automatically executed.The code,input,output,and execution of a smart contract are all open to users on the blockchain,which can facilitate cooperation between untrusted parties because the blockchain acts as a more trusted third party.In some cases,the blockchain validator may need to receive sensitive off-chain data to execute smart contracts.This sensitive information is also publicly visible,and this information will be stored on the blockchain forever.If the input data of the smart contract involves too much private information,this will limit the widespread use of the smart contract.On the one hand,wrong input information will lead to wrong contract execution results,so smart contracts require real off-chain input data for correct execution of contract business.On the other hand,users of smart contracts are afraid of exposing their sensitive information on the blockchain.Therefore,it is very important to ensure the privacy and authenticity of the input data of the smart contract at the same time,so that the users of the contract do not need to actively disclose sensitive information,and the smart contract can be executed correctly.This thesis first proposes an authenticated zero knowledge proof scheme zkDASNARK by extending the classical zk-SNARK scheme with data authentication.This scheme uses digital signatures to ensure the authenticity of data,and uses zkSNARK protects the privacy of the data,and the hash function associates the zeroknowledge proof with the digital signature,so that the authenticity and privacy can be guaranteed at the same time.Based on zk-DASNARK,this thesis then proposes a zero-knowledge authenticated off-chain data feed scheme zk-AuthFeed.As a general scheme,it can ensure the privacy and authenticity of any smart contract input data in the blockchain.This thesis also defines an ideal function to establish the security model of zk-AuthFeed,and then proves the security of the solution through simulation-based proofs.We has fully implemented the prototype of the zk-AuthFeed solution,and built a blockchain network consisting of 100 nodes on Aliyun.Based on the test network,this thesis conducted a full experiment on a medical insurance DApp,and considered 4 typical premium/reimbursement calculation models,including linear regression,logistic regression,support vector machine,and neural network.The results show that zk-AuthFeed has high efficiency:the key generation time does not exceed 12 seconds,the proof generation time does not exceed 4 seconds,the contract execution time is about 42 milliseconds,and the gas used of the contract is about 8×105 wei.