Article Published In Vol.9 (May-June 2021)

Assessing the Effectiveness of Multi-Factor Authentication in Strengthening Cybersecurity and Preventing Threats

Author : Charles Ubagaram and R. Pushpakumar

Cyberattacks have grown sophisticated in parallel with the advent of digital platforms, cloud computing, and remote work settings. Traditional password-based systems of authentication stand inadequate to protect sensitive data. Thus, Multi-factor Authentication (MFA) becomes a very vital security measure. Sadly, the implementation of MFA suffers from some hindrances such as user inconvenience, being attacked by hacktivists, and performance trade-offs. This research evaluates MFA in the light of strengthening cybersecurity and the prevention of threats using an Autoencoder Long Short-Term Memory (LSTM) model optimized with the African Buffalo Optimization (ABO) algorithm. The proposed solution applies deep learning technologies to the real-time detection of cyber threats and therefore secure. Data preprocessing using Min-Max Normalization increases the efficiency of the model, while Autoencoder LSTM sees to it that the detection of anomalies in network activities is done with precision. On top of that, the performance of the model is optimized further through the ABO by improving detection accuracy and decreasing the false positive rate. From the results, it shows that the proposed method achieves around 96% in terms of accuracy, thus improving dramatically the detection of threats as compared to traditional methods. Security is improved by 45%, with a reduction in unauthorized access risk, when compared with a system based on single-factor authentication. The realization of an AI-optimized MFA framework to counter cyber threats has been proven. Future work can be targeted towards integrating blockchain and behaviour analytics as stronger security measures.

Keywords: Multi-Factor Authentication, Cybersecurity, Cyber Threat Detection, Autoencoder LSTM, African Buffalo Optimization, Anomaly Detection.