Challenges of Event-based Streaming and Queuing as Data Exchange for Network Digital Twins

Network digital twins offer great potential for network management and monitoring, as they allow to use virtual representations of a real network, e.g., for risk-less tests and what-if analysis regarding fault, configuration, performance and security management. The digital twins take existing network emulation and simulation that is already established for these tasks a step further by allowing contiguous bidirectional state synchronization between the twins and the real network, ideally close to real-time. However, particularly in large networks, this holds significant challenges regarding the volume and velocity of the data to be exchanged between the twins and especially the real network. This paper presents an extension to the DigSiNet network digital twin environment to support message queues and event streaming platforms as scalable near real-time data exchange for the twins. Challenges for different common network management and monitoring data types are identified in an experimental setup and discussed regarding their application in use cases for network digital twins and the DigSiNet prototype. The implementation is provided as an open-source repository.