2026-04-13T06:38:37Zhttps://recercat.cat/oai/request

oai:recercat.cat:2117/4431572026-02-07T23:34:03Zcom_2072_1033col_2072_452951

00925njm 22002777a 4500 dc Niknahad, Hadi author 2025-07-04 As next-generation network infrastructures demand greater programmability, scalability, and autonomy, Intent-Based Networking (IBN) has emerged as a promising paradigm to simplify and automate service provisioning. IBN allows network operators to express service goals through human-readable, high-level intents that are then translated into low-level configurations by intelligent systems. However, in real-world multitenant networks, especially in the transport domain, handling conflicting or overlapping intents becomes a significant challenge, potentially leading to performance degradation or SLA violations. This thesis presents the design, implementation, and validation of a conflict management framework integrated into an enhanced IBN system, named Intent-Based Networking-Intent Management Entity (IBN-IME). The proposed architecture features a natural language processing-based intent interpreter, a semantic information model to align user expectations, and a fulfilment engine capable of autonomously detecting and resolving conflicts. Conflict detection follows a two-stage approach: intra-intent verification to ensure internal consistency, and inter-intent analysis to evaluate interactions between concurrent intents. A constraint evaluation module extracts and analyses SLA parameters such as bandwidth, latency, and packet loss, comparing them against operational limits. To address detected conflicts, the system applies a hybrid resolution strategy that combines priority-based handling, weighted fair sharing, and SLA softening mechanisms. The complete solution is implemented and integrated with the open source ETSI-compliant TeraFlowSDN controller. Experimental evaluation using a 14-node emulated testbed and more than 1,000 intents demonstrates that the framework achieves real-time conflict management while maintaining SLA adherence and supporting scalable, intent-driven orchestration in dynamic SDN environments. 9 - Indústria, Innovació i Infraestructura Àrees temàtiques de la UPC::Enginyeria de la telecomunicació Network analysis (Planning) Computer networks Artificial intelligence Intent-based networks Conflict management NFV SDN Netwotk Anàlisi de xarxes (Planificació) Xarxes d'ordinadors Intel·ligència artificial Intent-based conflict management in transport network services