Time-triggered traffic routing and scheduling optimization method based on cyclic queuing and forwarding
The Time-Sensitive Network(TSN)based on traditional Ethernet provides deterministic and reliable communication guarantees for time-sensitive applications.The TSN working group has proposed various shaping mechanisms to achieve traffic scheduling,among which the cyclic queuing forwarding mechanism based on a ping-pong queue simplifies gate configuration and enables bounded delay transmission of traffic.Existing traffic scheduling based on CQF mechanisms ignore the impact of routing policies on scheduling optimization.To address the queue congestion caused by the Shortest Path First(SPF)algorithm,this paper proposes a path selection algorithm based on a hybrid particle swarm algorithm.The algorithm considers minimizing path length and link load balancing to determine the optimal path while proposing a sorting method based on the QuickSort algorithm to improve the network's scheduling capability.Experimental simulations demonstrate that the algorithm can solve the scheduling problem of thousands of time-sensitive traffic flows in large-scale networks,and the successful rate of traffic scheduling can be increased by about 20% within the feasible execution time range.