Conference Paper

iREX: Efficient Inter-Domain QoS Policy Architecture.

Authors:
To read the full-text of this research, you can request a copy directly from the authors.

Abstract

Abstract, The inter-domain Resource Exchange (iREX) ar-chitecture uses economic market mechanisms to automate the deployment of end to end (E2E) inter-domain (ID) quality of service (QoS) policy among resource consumer and resource provider Internet Service Providers (ISPs). Previous simulation results have shown that iREX allows more coexisting ID policy deployments with less network congestion when compared to the existing method. In this paper we explore iREX's network load distribution efficiency limits by comparing iREX's performance to a lower bound for network congestion. We present an ana-lytical model of iREX in terms of a min-cost flow problem, and numerical results of efficiency loss between iREX simulations and derived optimal solutions based on multi-commodity flow optimization models. Our results show that for nominal to high traffic loads of 50% or more, iREX deviates a maximum of approximately 30% from the derived lower bound, while the current method deviates a maximum of 350%. Index Terms, inter-domain; QoS policy; resource allocation and management; network control by pricing; economics.

No full-text available

Request Full-text Paper PDF

To read the full-text of this research,
you can request a copy directly from the authors.

... iREX achieves this lowering of congestion by efficiently taking advantage of unused resources along multiple routes for the same source-destination domain pair. In [12], Yahaya, Harks and Suda explored the issue of iREX's efficiency further by addressing the question "how does iREX's performance compare to centralized optimal solutions?" ...
... In this paper, we detail and strengthen the work on the efficiency of iREX with respect to congestion started in [12] by incorporating an analytical analysis of the iREX architecture based on an online multi-commodity routing model. In this model, we prove upper bounds on the efficiency loss of the iREX architecture for arbitrary demands and arbitrary network topologies under the simplifying assumptions that: (i) reservations are non-expiring, and (ii) capacity constraints are inactive. ...
... Our main result with the numerical methodology shows that even though in the iREX architecture demands are routed selfishly and online (i.e. the chosen path is the cheapest and future demands are not known until the time they are routed), the efficiency loss (i.e. the percentage difference of the congestion experienced between iREX and the lower bound) is very low. The numerical results have partially appeared in [12]. ...
Article
The inter-domain resource exchange (iREX) architecture uses economic market mechanisms to automate the ad-hoc negotiation and deployment of end to end inter-domain quality of service policy among resource consumer and resource provider . In this paper, we explore iREX's network load distribution by comparing its performance to a lower bound for network congestion in two ways. We first present an analytical model of iREX in terms of an online algorithm and analyze its efficiency via competitive analysis. Our main result shows that the efficiency loss of iREX with respect to monetary cost is upper-bounded by a factor of 8 K/2 K+1, where K s the number of deployments, provided affine linear price functions are used. When the price functions are used to model congestion in the network, this result implies upper bounds on the efficiency loss of iREX with respect to network congestion. We then complement the analytical model with a numerical study using simulations.with optimal solutions derived from unsplittable and splittable multi-commodity flow optimization models. Our numerical results show that for nominal to high traffic loads of 40% or more, iREX deviates a maximum of about 20% from the lower bound, while the current method deviates a maximum of 300%.
... Bandwidth switching exchanges like Tradingcom Europe [27] are centralized services that operate similar to stock exchanges where ISPs trade excess capacity – iREX is a fully distributed architecture that can be used for similar purposes, but without the use of any centralized entity. Preliminary work on iREX was previously published in [1], work defining the initial iREX architecture was previously published in [2], and work exploring the efficiency of iREX was previously published in [3]. ...
Conference Paper
The inter-domain Resource Exchange (iREX) architecture uses economic market mechanisms to automate the deployment of end-to-end (E2E) inter-domain (ID) quality of service (QoS) policy among resource consumer and resource provider Internet Service Providers (ISPs). In iREX, each policy reservation is deployed on a single E2E ID path made up of the most "desirable" (i.e. cheapest and least congested) ISP resources. To accommodate ISPs that prefer redundancy when deploying ID QoS policy, in this paper we introduce an extension to the iREX architecture that gives an originating ISP a multi- path option (MPO) when deploying a reservation. MPO takes an initiating ISP's preference for redundancy and provides information about the available path options to achieve this preference in a distributed manner. Our simulation results show that while providing redundancy to the originating ISP using MPO does increase its resource costs in accordance to an ISP's preference, it only marginally increases overhead, and does not affect overall network performance - in fact the use of MPO lowers congestion.
Conference Paper
Full-text available
Open shortest path first (OSPF) is the most commonly used intra-domain Internet routing protocol. Traffic flow is routed along shortest paths, splitting flow at nodes where several outgoing links are on shortest paths to the destination. The weights of the links, and thereby the shortest path routes, can be changed by the network operator. The weights could be set proportional to their physical distances, but often the main goal is to avoid congestion, i.e., overloading of links, and the standard heuristic recommended by Cisco is to make the weight of a link inversely proportional to its capacity. Our starting point was a proposed AT&T WorldNet backbone with demands projected from previous measurements. The desire was to optimize the weight setting based on the projected demands. We showed that optimizing the weight settings for a given set of demands is NP-hard, so we resorted to a local search heuristic. Surprisingly it turned out that for the proposed AT&T WorldNet backbone, we found weight settings that performed within a few percent from that of the optimal general routing where the flow for each demand is optimally distributed over all paths between source and destination. This contrasts the common belief that OSPF routing leads to congestion and it shows that for the network and demand matrix studied we cannot get a substantially better load balancing by switching to the proposed more flexible multi-protocol label switching (MPLS) technologies. Our techniques were also tested on synthetic internetworks, based on a model of Zegura et al., (1996), for which we did not always get quite as close to the optimal general routing
Article
Full-text available
The Border Gateway Protocol (BGP) is an inter-Autonomous System routing protocol. It is built on experience gained with EGP as defined in RFC 904 [1] and EGP usage in the NSFNET Backbone as described in RFC 1092 [2] and RFC 1093 [3]. The primary function of a BGP speaking system is to exchange network reachability information with other BGP systems. This network reachability information includes information on the list of Autonomous Systems (ASs) that reachability information traverses. This information is sufficient to construct a graph of AS connectivity from which routing loops may be pruned and some policy decisions at the AS level may be enforced.
Article
Full-text available
. This is a preliminary version of a paper prepared for the Tenth Michigan Public Utility Conference at Western Michigan University March 25--27, 1993. We describe the history, technology and cost structure of the Internet. We also describe a possible smart-market mechanism for pricing traffic on the Internet. Keywords. Networks, Internet, NREN. Address. Hal R. Varian, Jeffrey K. MacKie-Mason, Department of Economics, University of Michigan, Ann Arbor, MI 48109-1220. E-mail: jmm@umich.edu, halv@umich.edu. Some Economics of the Internet Jeffrey K. MacKie-Mason Hal R. Varian The High Performance Computing Act of 1991 established the National Research and Education Network (NREN). The NREN is sometimes thought of as the "successor" to the NSFNET, the so-called backbone of the Internet, and is hoped by some to serve as a model for a future National Public Network. It is widely expected that substantial public and private resources will be invested in the NREN and other high performance ...
Article
Full-text available
. LOQO is a system for solving smooth constrained optimization problems. The problems can be linear or nonlinear, convex or nonconvex, constrained or unconstrained. The only real restriction is that the functions defining the problem be smooth (at the points evaluated by the algorithm). If the problem is convex, LOQO finds a globally optimal solution. Otherwise, it finds a locally optimal solution near to a given starting point. This manual describes (1) how to install LOQO on your hardware, (2) how to use AMPL together with LOQO to solve general convex optimization problems, (3) how to use the subroutine library to formulate and solve convex optimization problems, and (4) how to formulate and solve linear and quadratic programs in MPS format. 1. INTRODUCTION LOQO is a system for solving smooth constrained optimization problems in the following form: minimize f(x) subject to a #h(x)# b l # x # u. Here, the variable x takes values in R n , l and u are given n-vectors, f is a ...
Thesis
Die Dissertation befaßt sich mit der Verwendung von Modellierungssprachen zur schnellen Umsetzung praktischer Probleme in mathematische Modelle. Anhand verschiedener Beispiele wird gezeigt, daß es heute möglich ist, die auf diese Weise erzeugten gemischt-ganzzahligen Probleme mit Standard-MIP-Lösern zu lösen und so Problemstellungen zu bearbeiten, die noch vor wenigen Jahren die Implementierung spezieller Branch-and-Cut Algorithmen benötigten. Im ersten Teil der Arbeit wird die Modellierungssprache Zimpl vorgestellt. Kapitel 2 enthält eine vollständige Beschreibung der Sprache. Im nachfolgenden Kapitel werden Details der Implementierung von Zimpl, sowohl aus theoretischer als auch aus praktischer Sicht beschrieben. Dabei wird besonders auf Methoden des Software-Engineering zur Fehlerentdeckung und Vermeidung eingegangen. Im zweiten Teil werden mehrere Projekte beschrieben, bei denen die angeführte Methodik eingesetzt wurde. Kapitel 4 beschreibt drei Standortplanungsprojekte aus dem Telekommunikationsbereich. Kapitel 5 geht auf Fragestellungen bei der UMTS-Planung ein. Es werden die Problemstellungen, Modellierungen und die Lösungen erörtert, dabei wird besonders auf die Abhängigkeit der Ergebnisse von den Eingangsdaten und auf unerwartete bzw. unerwünschte Lösungen eingegangen. Abschließend wird ein bekanntes schweres kombinatorisches Problem, das Steinerbaumpackungsproblem in Graphen, aufgegriffen. Es wird eine bekannte, aber bisher nicht genutze Modellierung verwendet, die das klassische Switchbox-Verdrahtungsproblem und das Via-Minimierungsproblem vereint. Es gelingt, alle in der Literatur bekannten, sowie einige neu erzeugte, größere Probleminstanzen zu lösen.
Conference Paper
This paper introduces the inter-domain Resource Exchange (iREX) architecture for the automated deployment of fault tolerant end to end (E2E) inter-domain (ID) quality of service (QoS) policy among resource user and resource provider Internet Service Providers (ISPs). iREX uses economics and fully distributed mechanisms to empower domains to self-manage the deployment of E2E ID QoS policy. In iREX, resource user domains select and reserve ID QoS resources to form E2E ID QoS policy while resource provider domains provide information about available ID QoS resources and support the deployment of policies. iREX promotes accountability by enabling the establishment of bilateral agreements between a resource user ISP and all resource provider ISPs, and promotes congestion- avoidance by enabling a distributed resource selection process that selects the least congested ID deployment path. With iREX, domains cooperate to deploy ID QoS policy while maintaining their autonomy at all times. Simulation results show that iREX blocks fewer reservations and causes less congestion than the existing method, and that iREX is able to recover from faults.
Article
We study a dynamic, usage- and congestion- dependent pricing system in conjunction with price-sensitive user adaptation of network usage. We first present a Resource Nego- tiation and Pricing (RNAP) protocol and architecture to enable users to select and dynamically re-negotiate network services. In the second part of the paper, we develop mechanisms within the RNAP architecture for the network to dynamically formulate prices and communicate pricing and charging information to the users. We then outline a general pricing strategy in this context. We discuss candidate algorithms by which applications (singly, or as part of a multi-application system) can adapt their rate and QoS requests, based on the user-perceived value of a given combination of transmission parameters. Finally, we present ex- perimental results to show that usage and congestion-dependent pricing can effectively reduce the blocking probability, and allow bandwidth to be shared fairly among applications, depending on the elasticity of their respective bandwidth requirements.
Article
This paper presents the design of a new Internet routing architecture (NIRA). In today's Internet, users can pick their own ISPs, but once the packets have entered the network, the users have no control over the overall routes their packets take. NIRA aims at providing end users the ability to choose the sequence of Internet service providers a packet traverses. User choice fosters competition, which imposes an economic discipline on the market, and fosters innovation and the introduction of new services.This paper explores various technical problems that would have to be solved to give users the ability to choose: how a user discovers routes and whether the dynamic conditions of the routes satisfy his requirements, how to efficiently represent routes, and how to properly compensate providers if a user chooses to use them. In particular, NIRA utilizes a hierarchical provider-rooted addressing scheme so that a common type of domain-level route can be efficiently represented by a pair of addresses. In NIRA, each user keeps track of the topology information on domains that provide transit service for him. A source retrieves the topology information of the destination on demand and combines this information with his own to discover end-to-end routes. This route discovery process ensures that each user does not need to know the complete topology of the Internet.
Article
We study the relation between the number of firms and price-cost margins under price competition with uncertainty about competitors' costs. We report an experiment in which two, three, and four identical firms repeatedly interact in this environment. In line with the theoretical prediction, market prices decrease with the number of firms, but on average stay above marginal costs. Pricing is more aggressive than in equilibrium. Absolute and relative surplus increases with the number of firms. Total surplus is close to the equilibrium level, because enhanced consumer surplus through lower prices is counteracted by occasional displacements of the most cost-efficient firm. (JEL C90, C72, D43, D83, L13) Copyright 2005, Oxford University Press.
Conference Paper
Quality-of-service (QoS) encompasses a wide range of non-functional characteristics, including reliability, security, and performance and is becoming increasingly important in business-critical distributed systems. Such systems are increasingly deployed in open networks, such as the Internet, where resource allocation and load varies highly. To meet QoS requirements in such an environment, systems need to be able to dynamically (re)configure and adapt to changing QoS conditions. This paper describes a model for a QoS negotiation mechanism. The model allows clients and servers in distributed object systems to negotiate for QoS agreements involving multiple QoS categories, such as performance, reliability, security, etc. The model includes both a protocol that enables agents to negotiate and a technique for calculating the worth of alternatives. The protocol handles negotiations involving multiple offers and counter offers. It evaluates an offer based on the absolute requirements for QoS levels and their relative worth to the negotiating systems. The protocol and the worth calculations are part of a QoS negotiation model that involves simultaneous negotiation over multiple QoS categories. We describe how negotiating applications communicate and evaluate offers. In addition, we describe the functions the negotiation mechanism requires from its environment. The protocol has been formally specified and simulated. We are currently investigating implementation approaches and prototypes for the full as well as simplified versions of the described model
Article
Network delivery services providing "better-than-best-e#ort" service over the Internet are being studied, and are particularly necessary for multimedia applications. The selection and use of a specific delivery service involves negotiation between the user and the network; they agree upon specifications such as the type of service user packets will receive, the constraints the user tra#c must adhere to, and the price to be charged for the service. In this paper, we describe a protocol through which the user and the network (or two network domains) can negotiate network services. We refer to the protocol as a Resource Negotiation and Pricing protocol (RNAP). Through RNAP, the network service provider communicates availability of services and delivers price quotations and charging information to the user, and the user requests or re-negotiates services with desired specifications for one or more flows. RNAP protocol mechanisms are flexible enough to support multiple delivery service models, and allow dynamic re-negotiation of services during a session. Two di#erent network architectures are defined to support RNAP, a centralized architecture with a Network Resource Negotiator (NRN) administering each network domain, and a distributed architecture without any centralized controlling entity. Mechanisms are proposed for local price and charge computation, formulation of end-to-end prices and charges across multiple domains, and communication of this information through RNAP messages. Results of a prototype implementation are briefly described.
Article
As the Internet makes the transition from research testbed to commercial enterprise, the topic of pricing in computer networks has suddenly attracted great attention. Much of the discussion in the network design community and the popular press centers on the usage-based vs. flat pricing debate. The more academic literature has largely focused on devising optimal pricing policies; achieving optimal welfare requires charging marginal congestion costs for usage. In this paper we critique this optimality paradigm on three grounds: (1) marginal cost prices may not produce sufficient revenue to fully recover costs and so are perhaps of limited relevance, (2) congestion costs are inherently inaccessible to the network and so cannot reliably form the basis for pricing, and (3) there are other, more structural, goals besides optimality, and some of these goals are incompatible with the global uniformity required for optimal pricing schemes. For these reasons, we contend that the rese...
Article
Bandwidth brokers as proposed in the diffserv framework build on the process of setting up Service Level Agreements (SLA). This is a very static procedure, usually performed manually and based only on a simple description of the SLA. However, diffserv is suited to handle a more dynamic environment and to provide more than connectivity with its service classes. Our proposal enhances bandwidth brokers by including more significant information in SLAs such as the flow's destination network and pricing. In addition, we present a new per-hop behavior that supports flexible allocation of code points which is important to our approach. The system is targeted at large ISPs with good inter-connectivity. We implemented it on top of a flow-based simulation engine. The evaluation is based on parts of the current AS topology and characteristics of aggregated traffic. The results show an improvement of network utilization by up to 40% over a traditional, shortest-path routed inter-domain network for...
Article
The Internet continues to evolve as it reaches out to a wider user population. The recent introduction of user-friendly navigation and retrieval tools for the World Wide Web has triggered an unprecedented level of interest in the Internet among the media and the general public, as well as in the technical community. It seems inevitable that some changes or additions are needed in the control mechanisms used to allocate usage of Internet resources. In this paper, we argue that a feedback signal in the form of a variable price for network service is a workable tool to aid network operators in controlling Internet traffic. We suggest that these prices should vary dynamically based on the current utilization of network resources. We show how this responsive pricing puts control of network service back where it belongs: with theusers. 1. Introduction A communications network is as good, or as bad, as its users perceive it to be. Network performance should therefore be measured in terms of...
© 1-4244-0357-X/06/$20.00 2006 IEEE This full text paper was peer reviewed at the direction of IEEE Communications Society subject matter experts for publication in the IEEE GLOBECOM
  • Tradingcom Europe
Tradingcom Europe, " http://www.tradingcomeurope.com. " © 1-4244-0357-X/06/$20.00 2006 IEEE This full text paper was peer reviewed at the direction of IEEE Communications Society subject matter experts for publication in the IEEE GLOBECOM 2006 proceedings.
FRF.13 service level definitions implementation agreement
  • Frame Forum
Frame Relay Forum, " FRF.13 service level definitions implementation agreement, " August 1998.
GB917 Eva.Ver.1.5 SLA management handbook
TeleManagement Forum, " GB917 Eva.Ver.1.5 SLA management handbook, " June 2001.
  • Y Rekhter
  • T Li
Y. Rekhter and T. Li, " A border gateway protocol 4 (BGP-4), " IETF, RFC 1771, March 1995.
Pricing the internet