A CQF Scheduling Algorithm Based on ACO in Time-Sensitive Networking

The booming development of intelligent manufacturing has increasingly demanded scalability and flexibility of industrial networks. Dedicated industrial networks hinder industrial development due to rigid architecture and incompatible protocols, unable to support low-cost production adjustments and cross-unit collaboration. Time-Sensitive Networking (TSN) delivers device interconnection, flexible adaptation and low-latency transmission with standardized Ethernet, dynamic resource allocation and deterministic scheduling, and is regarded as the optimal solution. As one of the core scheduling mechanisms, Cyclic Queuing and Forwarding (CQF) has become the focus of current research due to its advantages of concise scheduling logic, low implementation cost, and predictable latency. Although existing heuristic scheduling algorithms for CQF can quickly provide feasible solutions, the quality of the solutions depends on their fitness to the problem, which may lead to local optima. In this study, we propose a CQF Scheduling Algorithm based on ACO (ACO-CQF). The algorithm utilizes a spatiotemporal 2-dimensional optimization framework based on coupled core decision variables, achieving joint optimization of multi-path selection and injection slot allocation through roulette wheel selection and pheromone guidance. The experimental results show that ACO-CQF outperforms traditional heuristic algorithms in terms of scheduling success rate and resource utilization rate, while the growth rate of its time consumption is within an acceptable range.