CAVP-IB-QoS: A Context-Aware, Verifiable, and Proactive Framework for Intent-Based Network QoS

Modern networks require robust Quality of Service (QoS) guarantees in dynamic environments, a challenge for traditional and even Software-Defined Networking (SDN) solutions. Intent-Based Networking (IBN) offers a solution by enabling users to express high-level intents. This paper introduces CAVP-IB-QoS, a novel context-aware, verifiable, and proactive framework for QoS management in IBNs. CAVP-IB-QoS features a hybrid intent model, context-aware policy synthesis, adaptive policy enforcement, and proactive optimization. This work focuses on evaluating the policy synthesis and adaptation through simulation. We compare CAVP-IB-QoS with Weighted Fair Queueing (WFQ), Priority Queueing, 1 and Static Allocation, using a Python-based simulator with a simplified network topology and dynamic traffic. Results demonstrate that CAVP-IB-QoS significantly outperforms the baselines in intent fulfillment rate, while maintaining comparable average latency. CAVP-IB-QoS achieved a 36% mean intent fulfillment rate, substantially higher than WFQ (12%), Priority Queueing (28%), and Static Allocation (0%). This is attributed to dynamic prioritization of flows based on user intents and adaptive bandwidth allocation. While the evaluation uses simplified routing and topology, the findings highlight the potential of CAVP-IB-QoS for robust QoS management in IBNs. Formal verification and digital twin-based optimization (PQDT) are part of the complete CAVP-IB-QoS design, but are left for future work.