Elsevier

Computer Communications

Volume 30, Issue 6, 26 March 2007, Pages 1301-1314
Computer Communications

An efficient technique for a series of virtual topology reconfigurations in WDM optical networks

https://doi.org/10.1016/j.comcom.2006.12.013Get rights and content

Abstract

This paper studies a series of reconfiguration processes corresponding to a series of traffic demand changes in WDM optical networks. This study describes the reconfiguration problem from two perspectives. First, the reconfiguration problem is a multi-optimization problem such that a single-objective optimization method could not be applied. Second, the reconfiguration problem consists of a series of reconfiguration processes for given traffic demand changes. The current virtual topology optimally generated for a given traffic demand change may lead to the costly disruption of traffic routing to adapt a new traffic demand change because it is independent of the next traffic demand change. Since the reconfiguration process that considers the only current traffic demand cannot guarantee the optimal average outcome in the long term, sequential decision-making is required to optimize the average outcome from the entire series of reconfiguration processes. The proposed reconfiguration technique is to minimize costly changes of lightpath routing changes in terms of network cost and minimize the average hop distance of traffic in terms of network performance simultaneously. Since the objective goal of maximizing network performance conflicts with that of minimizing network cost, there exists a set of non-dominated solutions called the Pareto front. The reconfiguration technique includes a reconfiguration process and a reconfiguration policy. The reconfiguration process finds a set of non-dominated solutions using the PEAP (Pareto Evolutionary Algorithm adapting the Penalty method) that optimizes objective functions by using the concept of Pareto optimal. Then, reconfiguration policy picks a solution from the set of non-dominated solutions using the MDA (Markov Decision Action). A case study based on simulation experiments is conducted to illustrate the application and efficiency of the proposed technique. It shows that our reconfiguration technique incorporating the PEAP and the MDA yields efficient performance in the entire series of reconfiguration processes.

Introduction

There are two topologies in a WDM (Wavelength Division Multiplexing) optical network, a physical topology and a virtual topology. The physical topology consists of optical fiber links and photonic nodes. The virtual topology consists of a set of lightpaths that carry optical signals from source nodes to destination nodes for given traffic demands. A process of rearranging the virtual topology to meet new traffic demands is called a reconfiguration process [1]. The reconfiguration process is one of the advantages in a WDM optical network. However, reconfigurations in the virtual topology interrupt the flow of costly traffic. The reconfiguration process of a virtual topology is a major task when new traffic demands are given in a WDM optical network. When the previous traffic demands are changed over a period of time, the optimal reconfiguration of a virtual topology is required to minimize network cost and maximize network performance. Since the reconfiguration is not a one-time operation, it will be activated whenever the current traffic demands are changed. The consequent reconfiguration problem is how and when to perform a reconfiguration process. A reconfiguration policy should be considered to control the reconfiguration process and generate an optimal virtual topology in the long term. The reconfiguration process and the reconfiguration policy are challenging problems in WDM optical networks.

This paper is organized as follows. Section 2 presents reconfiguration process methodology available in literature. Section 3 presents the problem formulation for the design of proposed reconfiguration technique. In Section 4, we present our reconfiguration technique that describes the PEAP (Pareto Evolutionary Algorithm adapting the Penalty method) used for the reconfiguration process and the MDA (Markov Decision Action) used for the reconfiguration policy. In Section 5, the proposed reconfiguration technique is evaluated in terms of two significant, optimal objective goals: the average hop distance of traffic in terms of network performance and the minimal number of lightpath routing changes in terms of network cost. Section 6 concludes this paper.

Section snippets

Background and rationale

We discuss the existing solutions to the subproblems of reconfigurations in terms of optical network topologies, reconfiguration process, and reconfiguration policies. Previous research has been studied in three different optical network topologies; a broadcast optical network, a ring optical network, and a wavelength-routed optical network. There are several heuristic approaches in the reconfiguration problem of broadcast optical networks. Ref. [1] exploits a sequence of branch exchange

Problem formulation

In this section, we give problem formulations for a series of reconfigurations in WDM optical networks.

Reconfiguration technique

When traffic demands occur, a virtual topology needs to be reconfigured to serve the traffic and to retain high performance. In this section, we propose a reconfiguration technique that minimizes costly changes in a virtual topology and maximizes network performance for a series of reconfiguration in the long term. The reconfiguration framework consists of two objective functions (AHT and NLC) described in Section 3.2, the PEAP in Section 4.1, and the MDA in Section 4.2. The reconfiguration

Experiments

Fig. 5 shows the 14-node NSFNET network topology used for the performance measurement of the proposed reconfiguration technique [17]. Here are our assumptions for the experimentation. Each link illustrated in Fig. 5 consists of two unidirectional links in opposite directions. We assume that each node is working as both an access node and a routing node. Each node is capable of grooming traffic at its edge and equipped with a wavelength converter. The lightpath capacity is OC-192. The total

Conclusion

We propose a reconfiguration technique adapting multi-objective optimization in WDM optical networks. The reconfiguration problem in WDM optical networks requires a process of multi-objective optimization because the objective of reconfiguration considers the network performance and the network cost simultaneously. In this paper, the AHT is exploited for the measurement of network performance and the NLC is exploited for the measurement of network cost. The proposed reconfiguration technique

Acknowledgement

This work has been supported by “Research Center for Future Logistics Information Technology” hosted by the Ministry of Education in Korea.

Sungwoo Tak is an assistant professor in the Department of Computer Science and Engineering at Pusan National University. He is a member of the Research Institute of Computer Information and Communication. He received a Ph.D. degree in Computer Science from the University of Missouri – Kansas City. He received a NSF-EpSCoR New Faculty Career Research Award in 2004. He has served as a TPC member for the ICCCN 2004/2005/2006 and AINA2005. His research interests include Computer Networks, Software

References (19)

  • K. Lee et al.

    On the reconfigurability of single-hub WDM ring networks

    IEEE/ACM Trans. Network

    (2003)
  • H. Krishnamurthy et al.

    Restoration mechanisms for handling channel and link failures in optical WDM networks: tunable laser-based switch architectures and performance analysis

    Comput. Commun.

    (2005)
  • J.F.P. Labourdette et al.

    Branch-exchange sequences for reconfiguration of lightwave networks

    IEEE Trans. Commun.

    (1994)
  • A. Narula-Tam, E. Modiano, Dynamic load balancing for WDM-based packet networks, in: Proc. of IEEE INFOCOM 2000,...
  • P.H. Ernest, G. Mohan, V. Bharadwaj, An efficient algorithm for virtual topology reconfiguration in WDM optical ring...
  • I. Baldine, G.N. Rouskas, Dynamic load balancing in broadcast WDM networks with tuning latencies, in: Proc. of IEEE...
  • I. Alfouzan, A. Jayasumana, Dynamic reconfiguration of wavelength-routed WDM networks, in: Proc. of 26th Annual IEEE...
  • D. Banerjee et al.

    Wavelength-routed optical networks. linear formulation, resource budget tradeoffs, and a reconfiguration study

    IEEE/ACM Trans. Network.

    (2000)
  • J. Zheng, B. Zhou, H.T. Mouftah, Design and reconfiguration of virtual private networks (VPNs) over all-optical WDM...
There are more references available in the full text version of this article.

Cited by (0)

Sungwoo Tak is an assistant professor in the Department of Computer Science and Engineering at Pusan National University. He is a member of the Research Institute of Computer Information and Communication. He received a Ph.D. degree in Computer Science from the University of Missouri – Kansas City. He received a NSF-EpSCoR New Faculty Career Research Award in 2004. He has served as a TPC member for the ICCCN 2004/2005/2006 and AINA2005. His research interests include Computer Networks, Software Architecture, WDM Optical Networks, Real-time Operating Systems, Embedded Systems, SoC (System on a Chip) Design, and Security.

Passakon Prathombutr was born on August 8th, 1968 in Nakhon Sawan, Thailand. He graduated from Chiang Mai University, Thailand with a B.S. degree in Physics (HONS) in 1990. After that he pursued a higher degree at Chulalongkorn Univesity, Bangkok, and graduated a M.S. degree in Computer Science in 1993. Then he became a researcher at National Electronics and Computer Technology Center (NECTEC), Ministry of Science and Technology, Thailand. In 1996, he received a government scholarship for a doctoral degree in data communication field in the United States. He attended Oklahoma State University and graduated in 1998 with the second M.S. degree in Compute Science. Then he attended the interdisciplinary Ph.D. program in Software architecture and Computer networking at the University of Missouri, Kansas City and graduated the Ph.D. in December 2003. At NECTEC, he was a director of Next Generation Internet Program and Vice President of Thailand IPv6 Association. Currently he serves as a director of the Intelligent Transport Systems Program. His research areas include the IPv6, ITS, Network Security and Optical Networking.

E.K. Park is a Professor of Computer Science at the University of Missouri at Kansas City. He received a Ph.D. degree in Computer Science from the Northwestern University. His research interests include software engineering, software architectures, software agents, distributed systems, object-oriented methodology, software tolerance and reliability, computer networks and management, optical networks, database/data mining, numerical computing, optimizations, and information/knowledge management. Currently, he is on an assignment serving as a Program Director, Division of Computing and Communications Foundations at US National Science Foundation.

View full text