Pragati Node Popularity (PNP) Approach to Identify Congestion Hot Spots in MPLS

Authors: E. Ramaraj, A. Padmapriya

Abstract:

In large Internet backbones, Service Providers typically have to explicitly manage the traffic flows in order to optimize the use of network resources. This process is often referred to as Traffic Engineering (TE). Common objectives of traffic engineering include balance traffic distribution across the network and avoiding congestion hot spots. Raj P H and SVK Raja designed the Bayesian network approach to identify congestion hors pots in MPLS. In this approach for every node in the network the Conditional Probability Distribution (CPD) is specified. Based on the CPD the congestion hot spots are identified. Then the traffic can be distributed so that no link in the network is either over utilized or under utilized. Although the Bayesian network approach has been implemented in operational networks, it has a number of well known scaling issues. This paper proposes a new approach, which we call the Pragati (means Progress) Node Popularity (PNP) approach to identify the congestion hot spots with the network topology alone. In the new Pragati Node Popularity approach, IP routing runs natively over the physical topology rather than depending on the CPD of each node as in Bayesian network. We first illustrate our approach with a simple network, then present a formal analysis of the Pragati Node Popularity approach. Our PNP approach shows that for any given network of Bayesian approach, it exactly identifies the same result with minimum efforts. We further extend the result to a more generic one: for any network topology and even though the network is loopy. A theoretical insight of our result is that the optimal routing is always shortest path routing with respect to some considerations of hot spots in the networks.

Keywords: Conditional Probability Distribution, Congestion hotspots, Operational Networks, Traffic Engineering.

Digital Object Identifier (DOI): doi.org/10.5281/zenodo.1331637

References:

[1] D. Awduche et al, "Overview and principles of Internet Traffic Engineering" IETF, Internet draft, draft-ietf-tewg-principles-02.txt, Jan 2002.
[2] D. Awduche, J.Maledm, J. Agogbua, M.O-Dell and J. McManus, "Requirements for traffic Engineering over MPLS", IETF RFC 2702, Sep 1999.
[3] ATM Forums, Private-Network-Network-Interface Specification Version 1.0 (PNNI 1.0), International Standart-of-pnni-oo55.000, March 1996.
[4] Yufei Wang, Zheng Wang, Leah Zhang, "Internet traffic engineering without full mesh Overlaying", Infocom 2001.
[5] Raj P.H and SVK Raja, Identifying Congestion hot spots in MPLS using Bayesian network", Asia J.Inform. Technol., 6 : 854-858, 2007.
[6] F. Jensen, "An Introduction to Bayesian Networks", UCL Press, 1996.
[7] J. Moy, "OSPF version 2", Internet rtc, IETF May 1998, Technical report RFC 2328.
[8] Yufei Wang and Zheng Wang, Explicit routing algorithm for Internet Traffic Engineering", Proceeding of ICCCN-99, Sep 1999.
[9] C. Villamizer, "OSPF optimized multipoint", Internet draft draft-ietfospf- omp-00.txt, IETF, March 1998.