No abstract

Current infrastructures are reaching the point where existing networking methods are unable to cope with the exponential growth of traffic and Quality of Service (QoS) requirements. New techniques are necessary to keep pace. One such technique, Software-Defined Networking (SDN) uses a central controller to program many individual network devices. However, SDN uses heuristic algorithms that do not always select the optimal path. This paper looked at creating three Q-Routing algorithms leveraging SDN and Mesh network topologies. Two algorithms used one network metric each (Latency and Bandwidth) and the third used multiple metrics. Results showed that the single metric Q-Routing algorithms on average performed as well as the K-Shortest Path versions while Q-Routing with multiple network metrics failed to match K-Shortest Path (different combination of metrics means these algorithms are not comparable). Results also showed that Q-Routing was able to calculate paths faster than K-Shortest Path in both static and dynamic networks.

As Internet usage increases, new, smarter, networking methods are required to enhance or maintain Quality of Service (QoS). One method, Software-Defined Networking (SDN) offers many advantages by separating the Forwarding and Control Planes. However, heuristic routing algorithms employed by SDN, such as Shortest Path, are not always suited for QoS-based pathfinding. This paper introduces a new Q-Routing algorithm that separates training and pathfinding, utilising two network metrics -latency and bandwidth -instead of latency alone. Two versions of this algorithm are employed, a static and a dynamic version where additional re-training is undertaken to allow Q-Routing to adapt to changing network environments. Both are tested on different size mesh topologies. The results show that static and dynamic Q-Routing are faster at pathfinding compared to K-Shortest Path and on average, find equally good routes.

Q-routing is a promising approach to finding routes in software-defined networks. This paper describes (i) the extension of Q-routing to find paths satisfying multiple Quality of Service metrics such as bandwidth and latency, and (ii) the use of fuzzy matching to allow relaxation of the metrics. Preliminary simulations suggest that fuzzy relaxation of metrics increases the number of flows that can be handled under high network loads.