An ACO-based Elephant and Mice Flow Scheduling System in SDN

Software defined network with centralized control and flexibility of programming has been widely used. And congestion could occur easily in this kind of network owing to more elephant flows with increasing network requirements. To resolve the problem, this paper proposed an ant colony optimization algorithm based on flow scheduling system (TSACO). The system accurately detects elephant flows with OpenFlow and sFlow, computes paths for elephant flows by an adaptive multi-path calculation algorithm, transmits mice flows through the paths in path database. The results show the proposed system can eliminate the forwarding conflicts between elephant and mice flows, ensure real-time transmission of mice flows and avoid splitting elephant flows as far as possible.