Energy-Aware Mapping and Scheduling of Network Flows With Deadlines on VNFs
Chadi M. Assi, Sara Ayoubi, Nicolas El Khoury, Long Qu · IEEE Transactions on Green Communications and Networking · 2018
Hardware middleboxes are critical elements in today's networks. Despite their important roles, they are accompanied by several problems, namely, their lack of flexibility, high capital and operational expenditures, and power consumption. Owing to the recent advances in virtualization, network function virtualization promises to address these problems, through replacing hardware middleboxes by software-based entities which can run on commodity hardware. These virtual network functions (VNFs) promise to alleviate the numerous disadvantages brought by their hardware counterparts. One of these most serious issues is the steadily increasing power consumption. In order to further optimize the power consumption, an efficient framework capable of mapping and scheduling traffic on these VNFs is needed. Such a framework allows to optimally assign and schedule the flows to be serviced, and place the unused servers in energy saving modes. In this paper, we assume VNFs are already placed on physical machines and consider traffic flows with deadlines. We focus on the problem of assigning and scheduling flows to VNFs in the most energy efficient manner. We formulate this problem mathematically and, owing to its complexity, present an efficient algorithmic method for solving it. We compare our heuristic with two other approaches, one of which aims to minimize the makespan and the other to minimize number of servers used. We show that our heuristic combines the advantages of both approaches and generates better results by consuming up to 31.3% and 46.1% less energy than other two approaches.