Abstract

The ability to store data on a cloud server for scalable services was made possible by cloud computing. To collect cloud user data from many locations using wireless means, IoT includes dispersed sensor nodes. However, the authentication time wasn't reduced, and the accuracy of the authentication did not increase. An IoT framework with three-layer authentication known as the Adaptive Elgamal Cryptosystem Algorithm (AECA) Three-Layer Authentication Method is proposed to overcome these problems. Initially, we collected a Kaggle dataset. The dataset's target parameter includes details on ten thousand bank clients. Of these, 2037 samples belonged to the class of negativity, and 7963 samples belonged to the category of the positive category. The AECA Method aims to facilitate efficient IoT financial transactions with a higher level of confidentiality. It involves a cloud user as the buyer and a cloud server as the seller, with the Cloud Server generating public and private keys during the key generation phase. When a user needs to make a financial transaction, they use a key pair to log in and send a demand message to customer support. The cloud manipulator is considered and authorized when the user's password and fingerprint match the computer system's password. If not, the transaction is rejected. The study encryption time (13.64), decryption time (15.28), Key generation time (347.21) and security (94), considering variations in financial data entries and cloud users. By using the AECA Method, this strategy makes it possible to execute financial transactions quickly and securely.