Research On Multimodal Identity Authentication Technology Based On Blockchain

From mobile phone unlocking,community access control to airport security check,high-speed rail entry,biometric features such as fingerprints,faces,and irises that mark personal identity have become digital IDs for people to enter the Internet world.However,due to the complex and changeable detection conditions,authentication environment,and spoofing attacks,the accuracy,security,and stability of authentication based on single-modal biometrics need to be improved.Multi-modal uses multiple biometrics to overcome the limitations of single-modal authentication,and can make identity authentication systems safer and more robust.There are two ways to store identity data.One is to store all the collected biometrics on the central server,which has the risk of centralized leakage of user biometrics.Another way is to use a distributed system that the number of entities involved is huge and the degree of trust is low,and the security of biometric templates need to be strengthened.Coupled with the uniqueness of biometrics,the protection of biometric templates becomes particularly important,so it has become a current research hotspot.On this background,starting from the privacy of biometric templates,this paper uses fuzzy extraction technology and blockchain technology to encrypt and store biometrics in a decentralized manner.The main contents of the paper are as follows:(1)This paper expounds biometric authentication technology,biometric template protection technology and blockchain technology,and proposes a fuzzy extraction-based multi-tasking method for the problem that the template is lost and difficult to reuse when biometrics are stored in plaintext.Modal biometric template protection method.The designed fuzzy extractor FECNN is based on the convolutional neural network.The convolutional neural network is used as the feature extraction module,and FECNN is used to generate a random string as the key of the user's biometric template to mask the real feature value and achieve the effect of confidentiality.Through experiments,it is proved that the fuzzy extractor FECNN can correctly reconstruct the random key,and achieve effective privacy protection for the biometric template under the condition of ensuring the accuracy.(2)In view of the data security problems brought by the centralized storage of the current identity authentication mechanism,this paper designs a user identity chain User Chain according to the characteristics of blockchain decentralization and multi-party consensus,and stores the random key obtained by the fuzzy extractor.On the user identity chain,the PBFT consensus algorithm and smart contracts are used to realize on-chain registration and identity authentication.Such a scheme enables dual protection of biometrics and improved performance.The security of the scheme is proved by security analysis,and the usability of the scheme is also proved by experiments and performance analysis.