DP-DT: Data Plane Digital Twin Architecture to Handle Conflicts among SDN Applications

The evolution of networks demands increasingly stringent performance requirements from users, driven by the deployment of innovative services such as augmented reality and self-driving cars. To meet these demands, network management and control tasks require heightened flexibility and responsiveness. Network programmability paradigms, particularly Software Defined Networking (SDN), have emerged as a solution, facilitating the integration of Artificial Intelligence and the deployment of new functionalities. However, The proliferation of heterogeneous control agents in SDN environments can lead to conflicts, resulting in network instability, such as Service Level Agreement (SLA) violations, and inefficient resource usage. Thanks to its ability to enable the execution of what-if analysis, the Digital Twin (DT) technology offers a promising approach to handle conflicts among decisions made by different SDN applications. In this work, we propose a framework that integrates DTs of the network data plane, introducing a new architecture to support synchronization and communication within the ecosystem. Furthermore, a discussion on its integration on the classical SDN workflow is discussed. Additionally, a case study focusing on Industry 5.0 is presented to illustrate the implementation of the data plane DT, demonstrate the effectiveness of the framework in conflict resolution.