Optimization of common pool resource sharing in multidomain IP-over-WDM networks D. Staessens a,⇑ , D. Colle a , M. Pickavet a , A. Nowé b , K. Steenhaut c , P. Demeester a a Ghent University-IBBT, Department of Information Technology (INTEC), G. Crommenlaan 8, bus 201, B-9050 Ghent, Belgium b Vrije Universiteit Brussel, DINF Department, COMO Research Group, Pleinlaan 2, B-1050 Brussel, Belgium c Vrije Universiteit Brussel, ETRO Department, IRIS Research Group, Pleinlaan 2, B-1050 Brussel, Belgium article info Article history: Available online 1 July 2011 Keywords: Multilayer networks Multidomain networks Protection Availability abstract Carrier networks are gradually adopting a network model that consists of MPLS capable routers and OXCs interconnected by high bandwidth WDM links for transporting IP and Ethernet traffic. Globalization drives the quest for end-to-end QoS guarantees over different carrier networks. Recovery mechanisms are crucial to reach the high availability requirements of critical services. In this paper we investigate how to enable high availability services spanning multiple networks, using failure protection techniques, whilst obeying administrative constraints. We show how different multidomain schemes compare in terms of availability. A constructive proof is given of the applicability of our approach, which leads to a heuristic solution. We compare this heuristic solution to an ILP model. The heuristic finds optimum solutions in most cases, but for individual cases and in an operational scenario, the ILP shows there is still room for improvement. Ó 2011 Elsevier B.V. All rights reserved. 1. Introduction Communications services are playing a vital role in modern pri- vate, corporate and institutional life. This prevalent role is expected to continue to grow in importance for years to come. From the corporate and institutional point of view, strategic corpo- rate functions become more dependent on communications be- tween different offices and sites where even minor service interruptions can result in huge production delays and revenue loss. The transport networks that these businesses are relying on are also evolving to meet future requirements. Optical technologies such as wavelength division multiplexing (WDM) have drastically increased the bandwidth capacity at a low cost, expanding the ser- vice possibilities for these networks. On the other side, this cost- effectiveness has driven the competition between operators to the point that there is little revenue to be made from the classical phone/fax/data service. Looking to increase revenue, these opera- tors are now exploiting the capacity to introduce high bandwidth services such as broadband Internet access and digital television. This trend fuels the quest for a converged network architecture, able to run all voice, data and multimedia services, commonly called triple-play. The scalability and robustness of the Internet protocol (IP) suite are the main reasons for its success, therefore IP is the network layer protocol of choice for these future networks. Flexibility was an issue in IP, but recent developments with multi- protocol label switching (MPLS), have introduced very powerful traffic engineering extensions that can be used in IP networks [1]. The cost-effectiveness of IP networks for data communications and telephony through voice over IP (VoIP) solutions attracts a lot of attention from the corporate business community. While widely used for local area networks (LAN) at a single location, hav- ing a corporate IP network with internal VoIP spanning different sites is arguably the cheapest solution for handling internal data and voice traffic. To cope with this, most operators are offering VPN services over their network, subject to a service level agree- ment (SLA). These SLAs guarantee a certain quality of service (QoS) for the connections, but currently there are little guarantees for VPN services spanning multiple domains (such as the public Internet). 1.1. Network scenario A multidomain network consists of different independently operated subnetworks, called domains or autonomous systems. An autonomous system (AS) is defined as a set of routers under a single technical administration, using some interior gateway pro- tocol(s) (IGPs) and common metrics to route packets within the AS, and using an exterior gateway protocol to route packets to 0140-3664/$ - see front matter Ó 2011 Elsevier B.V. All rights reserved. doi:10.1016/j.comcom.2011.06.012 ⇑ Corresponding author. Tel.: +32 93314977. E-mail addresses: dimitri.staessens@intec.ugent.be (D. Staessens), didier.col- le@intec.ugent.be (D. Colle), mario.pickavet@intec.ugent.be (M. Pickavet), ann.no- we@vub.ac.be (A. Nowé), kris.steenhaut@vub.ac.be (K. Steenhaut), piet.demeester@intec.ugent.be (P. Demeester). Computer Communications 35 (2012) 531–540 Contents lists available at ScienceDirect Computer Communications journal homepage: www.elsevier.com/locate/comcom