Digi-infrastructure: Digital twin-enableD traffic shaping with low-latency for 6g smart cities

Digital twin (DT)-based smart cities are anticipated to achieve seamless integration between physical and digital objects to satisfy an enormous number of users across all domains. Therefore, the infrastructure of 6G smart cities has become an important topic. Many types and data priorities exist in 6G smart cities; therefore, data traffic management is challenging. Current solutions may face challenges adjusting to swiftly evolving network circumstances and the unexpected rise of time-sensitive data. They require flexibility to handle non-periodic, unforeseen, and time-sensitive traffic, such as mission-critical applications. While current research explores the combination of Time-Sensitive Networking (TSN) and 5G, the evolution to 6G also necessitates the integration of TSN and DT technology to achieve deterministic networking. Therefore, taking advantage of DT in data traffic management, we propose a DT-enabled traffic shaping architecture called Digi-infrastructure, consisting of an intelligent traffic shaper inspired by TSN. Our proposed shaper comprises two components: the first component is a frame classification method established on Deep Reinforcement Learning (DRL) to address the dynamic scheduling problem by minimising the end-to-end delay. The second component is an intelligent gate control mechanism that considers the time, queue status and specified transmission time of traffic classes according to priority based on latency requirements without using a gate control list or timing data gate control. Finally, our solution improves infrastructure connectivity, efficiency, and latency.