• Mininet Topology
• Network Optimization
• Network Performance
• OpenFlow
• Real-time Decisions
• Reinforcement Learning (RL)
• Ryu Controller
• Software Defined Networking (SDN)
• Traffic Routing

Software Defined Networking (SDN) is an emerging paradigm to manage and optimize Networks. The aim of this thesis is to optimize traffic routing in SDN environments by using RYU controller and Reinforcement Learning (RL Technique). To present our solution, we propose an optimal approach which is adding an RL agent in the Ryu controller which will decide the traffic routing in real-time. We expect to dynamically scale traffic flow with the delay feedback as well through basic RL algorithm, with OpenFlow. We present the benefits, performance, and efficiency of our approach via experiments on a mininet simulated topology. This study helps advanced SDN technologies by providing a method to actually deliver traffic optimization in dynamic network environments.