A time-triggered traffic scheduling method
Patent Information
- Application Number
- CN202611081460.0
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2026-07-21
- Publication Date
- 2026-10-16
AI Technical Summary
第一，解决现有技术中TT业务调度的路由规划与调度规划阶段相互割裂，路由阶段未考虑跨周期业务在时隙上的潜在冲突，导致下发给调度阶段的路由方案存在高冲突隐患，压缩调度优化空间的问题
[0023]This application discloses a time-triggered service scheduling method, comprising: defining the network topology as a directed graph and numbering the nodes and links; obtaining all service parameters to be scheduled; parsing the network topology and service parameters, and prioritizing the TT services based on the network topology betweenness coefficient; generating multiple candidate path sets for each TT service according to the ranking results using the K-shortest path algorithm, and comprehensively evaluating the average link load, variance of link load, and average link overlap of each candidate path to select the optimal transmission path; updating the load information of each link in the network after determining the route to achieve global load balancing; initializing algorithm parameters, a tabu list, and the global optimal solution; encoding the priority vector of the TT services to be scheduled as individual sparrows, randomly generating an initial population and calculating the fitness value; and so on. The responsiveness value is the number of services that cannot find a valid time slot that meets the time slot conflict-free constraint after simulated back-to-back time slot allocation under the current scheduling order. A heuristic initial solution is constructed according to the period and frame length to replace the worst individual in the population. An iterative loop is entered, updating the global optimal solution in each iteration and storing it in the historical optimal database. High-quality solutions are injected from the historical optimal database to replace some of the worse individuals. Position updates and taboo checks are performed on the worst individuals until the termination condition is met, at which point the final globally optimal scheduling order and the corresponding time slot planning result are output. In the iterative loop of scheduling planning, the set of services that have continuously failed scheduling in the global optimal solution is counted every preset number of times. If it is not empty, the corresponding service is triggered to return to the routing planning step to reselect a suboptimal path, and the relevant taboo records are cleared before scheduling continues. The technical effects of this application include: First, optimizing the evaluation dimensions of routing planning, achieving global load balancing of links, effectively eliminating local resource bottlenecks, and improving network resource utilization; Second, abandoning a single fixed sorting rule, combining business characteristics to achieve flexible scheduling and sorting, with strong scenario adaptability and higher scheduling success rate under large-scale business; Third, optimizing the initialization method and search mechanism of the metaheuristic algorithm, improving the quality of the initial solution, accelerating the algorithm convergence speed, and improving the solution efficiency.