Adaptive Traffic Aware Domain Isolation Framework for Multi-Controller Software-Defined Networks
Software-Defined Networking (SDN) is a new technology that makes the network programmable by decoupling the control plane from the data plane. In large-scale SDN deployments, however, a single controller can be a performance bottleneck due to high number of flow requests, slow control-plane response time, and its inability to withstand failures. In contrast to single-controller SDN architectures, multi-controller SDN architectures reduce the number of controllers required and distribute control across multiple domains, making coordination among multiple controllers and the isolation of static domains inefficient when traffic is dynamic. Based on the above, this paper presents an Adaptive Traffic-Aware Domain Isolation Framework (AT-DI) for multi-controller SDN environments. The proposed framework combines traffic-aware domain monitoring, controller-load estimation, adaptive inter-domain coordination, and dynamic flow-handling decisions. Rather than using a single synchronization method for all inter-domain flows, AT-DI chooses the appropriate coordination mode based on traffic locality, controller stress, and the volume of inter-domain traffic. The proposed architecture has been tested with the Ryu controller in a Mininet Fat-Tree topology and compared with Single-Controller SDN, HyperFlow, ONOS, Kandoo, Rama, and the static Domain Isolation Multi-controller (DiM) architecture. Experimental results demonstrate that AT-DI increases throughput, decreases latency and round-trip time, reduces inter-controller control overhead, and provides balanced controller utilization under various traffic loads. The results validate the claim that adaptive domain isolation is a scalable and lightweight control-plane solution for dynamic multi-controller SDN deployments.