An Innovative Secure Multimodal Biometric Verification and Encryption Scheme Using Integrity Key Cryptography and Feature Level Fusion in Cloud Computing

This work introduces a Secure Multimodal Biometric Authentication and Encryption System using Integrity Key Cryptography and Feature-Level Fusion for cloud computing (MBIC-FF). The system brings together data from fingerprint, iris and finger vein biometrics using Fisher Vector and weighted Gaussian Mixture Model, making the data available at the feature level for integration. The FMMF approach produces a bio-integrity key used in a revised AES-GCM-IV scheme that greatly increases protection of data on the cloud. The performance is tested using False Acceptance Rate (FAR), False Rejection Rate (FRR), encryption time and decryption time. FMMF-based authentication achieved a much better average FAR of 0.15 and a much lower FRR of 2.1, surpassing all the single modalities. FMMF files performed encryption and decryption in 28 ms and 31 respectively. Impersonation, spoofing and inside as well as DoS attacks are successfully stopped by the system which keeps cloud-based biometric authentication safe, efficient, and private.