Risk-Adaptive Multi-Factor and Continuous Authentication Framework for Intelligent Identity Verification in Cloud Computing
Abstract
Cloud computing environments demand identity verification mechanisms that jointly satisfy strong security and unobtrusive usability, yet conventional multi-factor authentication (MFA) verifies identity only once, at session initiation, leaving a persistent vulnerability window for the remainder of an active session. This paper presents RA-MFA-Cloud, a risk-adaptive multi-factor continuous authentication framework for intelligent identity verification in cloud computing. The framework fuses knowledge-based, possessionbased, and biometric authentication factors with behavioral biometrics profiling inside a dynamic risk-scoring engine. Unlike static MFA schemes, RA-MFA-Cloud continuously monitors user sessions and adaptively adjusts authentication requirements from real-time risk assessments that incorporate contextual, behavioral, and temporal features. The risk model computes a weighted composite score from session anomaly, device trust, geolocation consistency, and behavioral deviation, and a cloud-specific contextual analysis module folds multitenant isolation indicators, service-level access patterns, and resource-consumption anomalies into the same pipeline - a combination absent from prior cloud authentication frameworks. Evaluated on 10,000 cloud sessions collected from a production multi-tenant environment, RA-MFA-Cloud achieves 98.7% authentication accuracy, a 99.2% impostor detection rate (IDR), a 0.3% false acceptance rate (FAR), and 23 ms average decision latency, outperforming static MFA and single-factor baselines across every metric while adding only 1 ms of latency over static two-factor authentication. Ablation experiments isolate the individual contribution of each framework component, confirming that risk-adaptive scoring and behavioral biometrics are the two largest contributors to overall accuracy. These results indicate that RA-MFA-Cloud offers a practical, deployable path toward continuous, risk-aware identity assurance for multi-tenant cloud platforms.