摘要
针对Internet核心网日益严重的能耗问题,建立了绿色虚拟拓扑设计(GVTD)问题的形式化模型,通过业务汇聚、按需配置网络资源、动态虚拟拓扑设计和多粒度睡眠机制降低网络能耗。提出了一种基于约束路由的启发式算法--CBR-GVTD算法,利用单跳路由和多跳路由相结合的方法构建网络虚拟拓扑,并通过基于约束的路由实现网络功耗与路由性能的折衷。模拟结果表明,CBR-GVTD算法可在接口平均利用率为80%~90%和最大路由跳数不超过5的条件下,最多可降低62%~90%的网络功耗。
To reduce the energy consumption of Internet core network, a formulation model for green virtual topology design (GVTD) problem was developed, which employs the following measures:traffic aggregation, network resources allocation on demand, virtual topology design dynamically and multi-granularity sleeping. A heuristic algorithm base on constraint-based routing, namely CBR-GVTD, was proposed, which combines the single-hop and multi-hop routing to establish virtual topology, and makes trade-offs between the network power consumption and routing performance via the constraint-based routing mechanism. The simulation results indicate that CBR-GVTD can achieve at most 62%~90%power saving with very high resource utilization (80%~90%for interface utilization) and excellent routing performance (not more than 5 for maximum routing hops).
出处
《通信学报》
EI
CSCD
北大核心
2014年第4期112-123,共12页
Journal on Communications
基金
国家自然科学基金资助项目(61332001
61272104
61073045)
四川省杰出青年基金资助项目(2010JQ0011)
中国科学院计算技术研究所计算机体系结构国家重点实验室开放课题基金资助项目(ICT-ARCH201003)~~
关键词
绿色网络
基于约束的路由
虚拟拓扑设计
多粒度睡眠
核心网能耗
green networking
constraint-based routing
virtual topology design
multi-granularity sleeping
energy con-sumption of core network