Long-Short Term Memory Network Based Model for Reverse Brute Force Attack Detection

Abstract

Reverse brute force attacks pose a significant threat to the security of online systems, where adversaries attempt to gain unauthorized access by systematically testing a multitude of username and password combinations against a single account. To address this challenge, the research presents an innovative Long-Short Term Memory Network based model designed to detect such attacks. The model utilizes LSTM algorithms to analyze login attempt patterns, identifying anomalies that may indicate reverse brute force attacks. By examining various factors like user login behavior, IP address, and time-based patterns, the model distinguishes legitimate access attempts from potential attacks with high accuracy. It incorporates real-time threat intelligence feeds and historical data analysis to continuously adapt and improve its detection capabilities. The model dynamically adjusts security parameters, enforces account lockouts, and communicates with firewall systems to block suspicious IP addresses, thus providing a proactive response to thwart attacks. The research evaluates the effectiveness of the AI model through simulated and real-world testing scenarios, demonstrating a significant reduction in false positives and successful prevention of reverse brute force attacks. Overall, the developed AI model offers a sophisticated and proactive solution to the evolving threat of reverse brute force attacks, contributing to the advancement of cybersecurity measures.