Authentication System using Secret Sharing

Abstract

Security using Authentication system is an important concern in the field of information technology. It is an important thing as per as concern to the ruling of internet over people today. The growth in the usage of internet has increased the demand for fast and accurate user identification and authentication. This New threats, risks and vulnerabilities emphasize the need of a strong authentication system. The cryptography is a secret sharing scheme where a secret data gets divided into number of pieces called shares and not a single share discloses any information about secret data. There are some automated methods to identify and verify the user based on the physiological characteristics. To deal with such methods, there is a technology called biometrics which measures and statistically analyses the biological data. The biometric samples which are stored in the database as a secret are unique for each user so that no one can predict those samples. A biometric authentication system provides automatic authentication of an individual on the basis of unique features or characteristics possessed by an individual. A cover image is fused with secret image; fused image is divided into n shares. k possible shares able to construct secret image. PSNR parameter are used for image quality The authentication system can be stronger using multiple factors for authentication process. The application like Aadhar Card uses more than one factor for authentication. There is some difficulty with authentication systems such as user privacy considerations in case of multiple biometric features, huge size databases and centralized database which may create security threats. To address such tribulations, the Authentication System using Secret Sharing is proposed, Secret sharing splits the centralized database across the different locations. This helps in reducing the database size and removal of threats in centralized database. Also user privacy is maintained due to the decentralized database.