MODELING OF OPTIMAL MULTI KEY HOMOMORPHIC ENCRYPTION WITH DEEP LEARNING BIOMETRIC BASED AUTHENTICATION SYSTEM FOR CLOUD COMPUTING

Abstract

More recently, cloud computing (CC) has gained considerable attention among research communities and business people. Inspite of the advantages of CC, security, and privacy remains a challenging problem. Therefore, biometric authentication systems have been employed and fingerprint is considered as widely employed to attain security. In addition, image encryption techniques can be used to encrypt the fingerprint biometric image to add an extra level of security. Based on these motivation, this study designs an optimal multikey homomorphic encryption (OMHE) with stacked autoencoder (SAE) based biometric authentication system for CC environment. The proposed OMHE-SAE model aims to encrypt the biometrics using OMHE technique and then verification takes place using SAE model. In addition, the OMHE technique involves the optimal key generation process using sandpiper optimization (SPO) algorithm to effectively choose the keys for encryption and decryption. Furthermore, the verification of decrypted biometrics takes place by the use of SAE model. A wide range of simulation analyses take place on benchmark datasets and the experimental outcomes portrayed the betterment of the OMHE-SAE More than cutting edge technology.