A Topology Routing Optimization Algorithm for SDN Control Plane
The impact of different deployment methods and control paths on network energy consumption in multi con-troller scenarios for large data center networks,the problem of load imbalance caused by distributed cluster is solved.A dual objective optimization of energy-saving routing and load balancing between controllers is proposed in this paper.A Topology Routing Optimization Algorithm for SDN controller plane is studied,which can optimize the network flow,select the data path,and select the controller and control path for the activated switch,utilizing DQN to search for learning abilities and enhance decision-making abilities in reinforcement learning,the optimal path can be obtained by training only some states,and the switches and ports that need to sleep can be obtained,so as to determine the energy-saving network topology,effectively reduce the network energy consumption,and realize the effect of control plane load balancing at the same time.In this paper,dqn algorithm is used to solve the energy-saving data path and control path.Finally,the perfor-mance of the proposed algorithm is verified by experiments.
software defined networkinghigh energy consumption routingtopology routingDQN
万方数据
维普
