Rate control-based framework and algorithm for optimal provisioning

  • Authors:

  • G. Sun;H. Yu;L. Li;V. Anand;H. Di

  • Affiliations:

  • School of Communication and Information Engineering, University of Electronic Science and Technology of China, Chengdu, China 611731;School of Communication and Information Engineering, University of Electronic Science and Technology of China, Chengdu, China 611731;School of Communication and Information Engineering, University of Electronic Science and Technology of China, Chengdu, China 611731;Department of Computer Science, The College at Brockport, State University of New York, Brockport, USA 14420;School of Communication and Information Engineering, University of Electronic Science and Technology of China, Chengdu, China 611731

  • Venue:
  • Photonic Network Communications
  • Year:
  • 2011

Quantified Score

Hi-index 0.00

Visualization

Abstract

Ideally, networks should be designed to accommodate a variety of different traffic types, while at the same time maximizing its efficiency and utility. Network utility maximization (NUM) serves as an effective approach for solving the problem of network resource allocation (NRA) in network analysis and design. In existing literature, the NUM model has been used to achieve optimal network resource allocation such that the network utility is maximized. This is important, since network resources are at premium with the exponential increase in Internet traffic. However, most research work considering network resource allocation does not take into consideration key issues, such as routing and delay. A good routing policy is the key to efficient network utility, and without considering the delay requirements of the different traffic, the network will fail to meet the user's quality of service (QoS) constraints. In this paper, we propose a new NUM framework that achieves improved network utility while taking into routing and delay requirements of the traffic. Then, we propose a decomposition technique-based algorithm, D-NUM, for solving this model efficiently. We compare our approach with existing approaches via simulations and show that our approach performs well.