AN OPEN ARCHITECTURE FOR REAL-TIME QOS MONITORING AND MOS ESTIMATION IN IP-BASED NETWORKS

Abstract

The increasing demand for real-time IP-based services, particularly Voice over IP (VoIP), necessitates advanced and adaptive methods for Quality of Service (QoS) monitoring. This paper presents a modular, open-source architecture for real-time QoS assessment and perceptual quality estimation using Mean Opinion Score (MOS) modeling. The proposed system integrates passive traffic analysis, statistical QoS metrics, and predictive MOS evaluation. A layered design allows separation of concerns and flexible deployment. The system supports real-time alerting, QoS-aware service management, and enables early detection of degradation using normalized jitter, delay, and packet loss.