当前互联网存在着资源高冗余和高能耗等问题,随着网络规模的扩大和联网设备数量的激增,这些问题日益严重.文中在绿色互联网的背景下,着眼于主干网的拓扑管理和能耗控制,首先构建了主干网的网络模型,给出了节点和链路的结构模型,分析了网元功耗并建立了功耗模型;然后,基于上述模型,提出了一种绿色互联网中面向节能的分布式拓扑管理机制(Distributed Topology Management scheme for Energy saving,DTME),利用VCG(Vickrey-Clarke-Groves)机制对节点和链路进行配置管理,通过信息感知、流量预测、分布式拓扑决策和休眠控制之间的协同进行分布式拓扑管理,实现节能目标;最后,为了验证DTME在拓扑管理和节能方面的适用性和有效性,将其与采用GRiDA(GReen Distributed Algorithm)的拓扑管理机制在网络功耗和网络性能(收敛时间、平均延迟、分组丢失率和防策略性)等方面进行了仿真对比实验和性能评价,结果表明DTME具有良好的分布式拓扑管理能力和节能潜力.
The current Internet is confronted with some severe problems, for example, high resource redundancy and excessive energy consumption. With the continuous expansion of network scale and the sharply increasing of the number of networking devices, these problems become more and more serious. In this paper, topology management and energy consumption control in the backbone networks are investigated in the context of green Internet. The backbone network model is built, and its node and link structure models are proposed. The power consumption of network element is analyzed and its model is devised. Based on the above models, a Distributed Topology Management scheme for Energy saving (DTME) in green Internet is devised. In DTME, node and link configurations are managed with VCG (Vickrey-Clarke-Groves) mechanism, and distributed topology management and energy saving are achieved via collaboration among information sensing, traffic prediction, distributed topology decision-making and sleeping control. To verify the applicability and effectiveness of DTME in terms of topology management and energy saving,simulated experiments and performance evaluations are carried out to compare DTME with the GRiDA (GReen Distributed Algorithm) based topology management mechanism on network power consumption and network performance (including convergence time, average delay, packet loss rate and strategy-proofness). Experiment results demonstrate that DTME has strong distributed topology management ability and good energy saving potential.