Large-Scale Deterministic Network Time Sensitive Traffic Scheduling Mechanism with Joint Routing and Queuing

Aiming at the deterministic delay problem of time-sensitive traffic in large-scale deterministic networks, a joint routing and queue scheduling mechanism for time-sensitive network traffic is proposed. In order to solve the defects of the current scheduling mechanism, the scheduling mechanism proposed in this paper designs the real-time online scheduling model of joint routing and circular forwarding queue through the system architecture of the circular queue forwarding (CQF) mechanism based on the distributed SDN architecture and time sensitive network (TSN). Based on the DDQN algorithm, the model comprehensively considers the network status such as node queue, link capacity, feasible path, congestion information, and other traffic factors such as different levels of bounded delay requirements to achieve optimal scheduling of time-sensitive flows, aiming to improve the scheduling success rate and reduce delay and packet loss. By comparing this mechanism with the shortest path algorithm based on CQF mechanism and the DRL path selection algorithm based on CQF mechanism, the simulation results show that the proposed mechanism can flexibly adjust the transmission scale of traffic according to the resource usage of the network in the large-scale scheduling scenario, effectively improve the scheduling performance, and reduce the overall end-to-end delay.