A Strategy of Selecting a Maximal Set of Sleep Links for Energy Saving in Core IP Networks

Background/Objectives: Nowadays the energy consumption of ICT networks is about 10% of the total worldwide power consumption and is predicted to increase remarkably in the near future. Methods/Statistical Analysis: We study energy saving methods for IP networks in a way to select qualified links and place them in sleep mode under network-level QoS constraints. An energy saving strategy is proposed which is based on the concept of delegation process with a user-specified condition for limiting the increase in hop count of paths. The main feature of our strategy is delegation process through which a maximal set of sleep links can be found under the user-specified constraint. Findings: For maximizing energy saving, we present two heuristic algorithms for link selection under QoS constraints, and evaluate the characteristics of the algorithms by a computer simulation. Our two algorithms provide three kinds of methods, namely, max_set, min_traffic and min_traffic. This paper shows that in terms of link saving efficiency, the max_set is ranked first, the mini_traffic second, and the min_path last. The link saving efficiency of the max_set reaches about 65 % under certain conditions. Also, the max-set method is shown to find a larger set of sleep links than previous works, enabling more energy saving. On the certain conditions, we confirm the increased mean hops of the path within 1.2 hops in case of using the max_set for selecting sleep link. This value does not have a significant impact on the real end-to-end delay. Application/Improvements: We expect that our algorithms will provide the ISP with a mechanism for achieving the required level of energy saving by adjusting the delay of paths.