摘要
为改善网络节点资源利用率,提出基于Dyna agent算法的控制器控制策略。首先对网络控制器部署建模;然后控制器利用环境反馈信息反作用于控制端网络流量控制策略抉择;最后分析触发学习机制后各节点转发状态的变化,重新设置节点选择权值。仿真结果表明,该控制策略相对于普通控制策略的网络可使目的节点处理能力更强,吞吐量增大,网络延迟更小。
A control strategy based on Dyna agent algorithm is proposed to improve resource utilization of network node. First step is to model the deployment of the network controller,then the controller do policy choice by using the environmental feedback which reacts to controller terminal network. Finally,analyze the changes of the node forwarding situation after learning this method,and reset the node option value. The simulation results show that this control strategy has better ability to deal with process nodes,larger throughput and smaller network delay compared with the general control strategy.
出处
《计算机应用与软件》
CSCD
2016年第11期90-92,152,共4页
Computer Applications and Software
基金
陕西省教育厅科学研究计划项目(07JK377)
