Secure Multimodal Biometric Authentication for Smart ATMs Integrating Multiple Bank Accounts Using KFROR-KECCAK and DP-K

Nowadays, each and every bank uses individual Automated Teller Machine (ATM) cards to maintain accounts, thus challenging customers to handle more cards and their passwords. Hence, integration of multiple bank accounts in a single ATM with enhanced security and user convenience is important. But none of the prevailing works concentrated on the computation time in the authentication framework for the smart ATM environment in one smart ATM, thus resulting in delays and inefficiencies. Initially, the account registration phase starts by registering multiple bank and personal details using the DP-K technique. Then, biometric data, such as face, palm, and voice, are collected and pre-processed. Afterward, the features are extracted from the facial images, the minutiae of palm images, and the spectrogram-analyzed voice data. Next, the extracted features and facial points are converted into hash codes and stored in the database. Meanwhile, the traditional PIN is generated using a traditional random binary key. Now, the login phase initiates to verify biometrics, which are stored in the database. Then, the PIN verification is done with the generated PIN based on the PoM binary OTP. Once the verification is done, bank accounts and transaction pages are displayed. Thus, the proposed framework outperforms the other prevailing approaches by attaining 5645 kbps throughput.