Authentication of a System Using Audio File

Traditional text-based passwords are increasingly vulnerable to security threats such as bruteforce attacks, keylogging, and shoulder surfing. Although graphical password systems have emerged to enhance security, they too are susceptible to observationbased attacks. To address these limitations, we propose a novel audio-based password authentication system that prioritizes both security and user convenience. In this system, users select specific click points within an audio file as their password, thereby eliminating visual cues and significantly reducing the risk of shoulder surfing. During registration, up to four audio click positions are chosen, encrypted using AES-128 and securely stored using Least Significant Bit (LSB) encoding. For successful authentication, users must reproduce the same sequence of click positions. To enhance accessibility and usability, the system integrates a forgot password feature. If a user fails to authenticate after three consecutive login attempts, the system automatically triggers a Time-based One-Time Password mechanism. An OTP is generated using the TOTP algorithm and sent to the user's registered email address. This OTP is valid for a limited time window and is securely stored on the server for verification. Upon successful OTP verification, the user is allowed to reset their password. This ensures a secure recovery process without compromising the system's integrity. Additionally, the system incorporates a virtual driver robot on the backend that assists users by automatically and securely inputting login credentials, enhancing usability while preserving data privacy. The use of persuasive security techniques encourages users to create less predictable audio patterns, thereby mitigating the risks of guessing and brute-force attacks. This multilayered authentication system includes individual registration and login modules for platforms such as Facebook, Amazon, and the KCG portal, ensuring versatility and a seamless user experience across different environments.