Abstract: Resource management in Software-Defined Networking (SDN) is an important aspect in this brand new approach of networking. Implementing an efficient resource management with high Quality-of-Service (QoS) requirements like low delay or high bandwidth is a big challenge. Dynamic and adaptive solutions can be benefit for improving the performance of these networks. In this paper, we propose a novel approach of using genetic algorithm with SDN architecture and FNSS toolchain for better optimization of adaptive resource management with testbeds on datacenter topologies with video streams run on VLC media player to evaluate the performance of our solution. We explained the testbeds on the test video (delay and bandwidth) by implementing the adaptive solution using Genetic Algorithm (GA) for adaptively selecting the capacity according to link delay which then reduces the overall timeof packet delivery and increases the bandwidth. The final results shows an improvementin reducing the average delay values of 94.4% when using OVS controller and 90.1% when using Floodlight controller, and an improvement in bandwidth of 950 Mbps- 1.82 Mbps to 37.075 Mbps- 42.9 Mbps in OVS controller and 952 Mbps- 1.7 Mbps to 22.97 Mbps- 27.45 Mbps in Floodlight controller.
Keywords: SDN, Adaptive Resource Management, Fast Network Simulation Setup (FNSS), OVS Controller, Floodlight Controller, Mininet, Genetic Algorithm, Datacenter topology.