Internet traffic engineering

The main goal of Internet traffic engineering is to efficiently optimize the performance of operational networks in order to avoid the well-known shortcomings of the typical destination-based IP routing. Traffic engineering attempts to reduce or even avoid congestion hot spots and to improve resource utilization across the backbone IP network. During the last years traffic engineering has become an inevitable tool concerning performance optimization in large Internet backbones. The core objective of this paper is to give an overview of the architectures and mechanisms for traffic engineering.     

Practical OSPF traffic engineering

Open Shortest Path First (OSPF) traffic engineering (TE) is intended to bring long-awaited traffic management capabilities into IP networks, which still rely on today's prevailing routing protocols: OSPF or IS-IS. In OSPF, traffic is forwarded along, and split equally between, equal cost shortest paths. In this letter, we formulate the basic requirements placed on a practical TE architecture built on top of OSPF and present a theoretical framework meeting these requirements of practicality. The main contribution of our work comes from the recognition that coupled with an instance of the maximum throughput problem there exists a related inverse shortest-path problem yielding optimal OSPF link weights.     

Multi-timescale Internet traffic engineering

The Internet is a collection of packet-based hop-by-hop routed networks. Internet traffic engineering is the process of allocating resources to meet the performance requirements of users and operators for their traffic. Current mechanisms for doing so, exemplified by TCP's congestion control or the variety of packet marking disciplines, concentrate on allocating resources on a per-packet basis or at data timescales. This article motivates the need for traffic engineering in the Internet at other timescales, namely control and management timescales, and presents three mechanisms for this. It also presents a scenario to show how these mechanisms increase the flexibility of operators' service offerings and potentially also ease problems of Internet management.     

流量工程中静态路由算法的研究

该文提出了一种应用于流量工程环境中的静态路由算法。考虑当前的网络资源情况,分优先级别在网络中计算并配置标记交换路径(Label Switched Path,LSP),当某一优先级有多条 LSP需要并行配置时,利用遗传算法搜索最优或较优的配置方案,使得网络的链路带宽使用率低于管理员定义的某个限定值,达到合理分布资源的目的。此外,提出了一种改进的 Dijkstra 算法计算 LSP的最短路径。     

MPLS advantages for traffic engineering

This article discusses the architectural aspects of MPLS which enable it to address IP traffic management. Specific MPLS architectural features discussed are separation of control and forwarding, the label stack, multiple control planes, and integrated IP and constraint-based routing. The article then discusses how these features address network scalability, simplify network service integration, offer integrated recovery, and simplify network management. Scalability is addressed through integrated routing enabling a natural assignment of traffic to the appropriate traffic engineering tunnels without requiring special mechanisms for loop prevention. Change is greatly reduced. The label stack enables an effective means for local tunnel repair providing fast restoration. Feedback through the routing system permits fast and intelligent reaction to topology changes. Service integration is simplified through a unified QoS paradigm which makes it simple for services to request QoS and have it mapped through to traffic engineering     

Provisioning QoS in inter-domain traffic engineering

This work describes an architectural framework that allows inter-domain Traffic Engineering Label Switched Paths (TE-LSPs) with guaranteed quality of service (QoS) to be setup. Such TE-LSPs, called EQ-links, are setup by coordinating path computation elements (PCEs) of neighboring autonomous systems (ASs) along a pre-determined inter-AS path, computed through cooperative interaction between pairs of neighboring ASs. After defining the architectural requirements for the framework, we describe and analyze the Inter-AS Path Computation Protocol (IA-PCP), which computes an interdomain path at the AS level, i.e., selecting a sequence of ASs to the destination, based on a loose source routing approach. The results of the IA-PCP computations are then fed to the PCEs for complete path computation. The proposed architecture has been actually implemented within the testbed of the EuQoS project, which is aimed at enabling end-to-end QoS in the Internet. We report results related to the setup time of EQ-links, measured in the pan-European testbed of the EuQoS project, showing that path computation and setup takes an affordable time overhead.     

Autonomic traffic engineering for network robustness

The continuously increasing complexity of communication networks and the increasing diversity and unpredictability of traffic demand has led to a consensus view that the automation of the management process is inevitable. Currently, network and service management techniques are mostly manual, requiring human intervention, and leading to slow response times, high costs, and customer dissatisfaction. In this paper we present AutoNet, a self-organizing management system for core networks where robustness to environmental changes, namely traffic shifts, topology changes, and community of interest is viewed as critical. A framework to design robust control strategies for autonomic networks is proposed. The requirements of the network are translated to graph-theoretic metrics and the management system attempts to automatically evolve to a stable and robust control point by optimizing these metrics. The management approach is inspired by ideas from evolutionary science where a metric, network criticality, measures the survival value or robustness of a particular network configuration. In our system, network criticality is a measure of the robustness of the network to environmental changes. The control system is designed to direct the evolution of the system state in the direction of increasing robustness. As an application of our framework, we propose a traffic engineering method in which different paths are ranked based on their robustness measure, and the best path is selected to route the flow. The choice of the path is in the direction of preserving the robustness of the network to the unforeseen changes in topology and traffic demands. Furthermore, we develop a method for capacity assignment to optimize the robustness of the network.     

Creating multipoint-to-point LSPs for traffic engineering

Traffic engineering enhances an ISP's capability to manage and utilize its resources effectively. MPLS has emerged as an efficient packet forwarding tool that gives a significant boost to the traffic engineering capabilities of an ISP. A fundamental problem in MPLS is to reduce label space usage by label switched paths while meeting the requirements of the flows traversing the network. Using multipoint-to-point LSP trees has been proposed as one of the techniques to reduce label space usage. We look at the problem of creating multipoint-to-point LSPs given a set of precomputed point-to-point LSPs. We propose a heuristic for multipoint-to-point LSP creation and show its effectiveness.     

Multilayer traffic engineering for energy efficiency

Automatically switched multilayer IP-over-optical networks offer extensive flexibility in adapting the network to offered IP/MPLS traffic. Multilayer traffic engineering (MLTE) takes advantage of this through online IP logical topology reconfiguration in addition to the more traditional rerouting. The main goal of MLTE is to optimize toward resource usage, bandwidth throughput and QoS performance. However, energy efficiency of ICT infrastructure and the network in particular more recently have become an important aspect as well. In this article, we will look how MLTE helps in improving network energy efficiency. For this we will explain how optimization toward power requirement relates to the traditional resource usage minimization objective, and how power requirement in the network can be modeled for the MLTE algorithm. We will discuss two cases where the merit of MLTE for energy efficiency is discussed. Firstly, we will examine the interaction of MLTE with hardware-based energy efficiency optimization techniques; for this we look at scaling back power requirements through the use of better chip technology, but also decreasing idle-power requirement only, using improved chip architecture. Secondly, as MLTE allows for fast responses to changing traffic, we will see how link switch-off during off-peak hours offers a straightforward option to reduce energy needs.     

Multilayer traffic engineering for multiservice environments

Multilayer traffic engineering (MLTE) serves to provide cross-layer online network optimization techniques to cope with rapid variations and short-term evolutions in traffic patterns. MLTE extends traffic engineering as it exists in IP/MPLS-based technology toward the multilayer IP/MPLS-over-optical transport network. In addition to the IP/MPLS traffic routing, MLTE exposes much larger adaptation flexibility by building on next-generation automatic switched optical transport networks. These offer fast setup and teardown of end-to-end multi-hop optical connections (lightpaths), which are offered to the IP/MPLS layer as dynamically provisioned capacity. This dynamic nature leads to an IP/MPLS logical topology that can be reconfigured on the fly, and IP/MPLS link capacity that can be up- or downgraded as client traffic demand varies. These MLTE techniques are generally used to increase perceived network performance in terms of throughput or QoS. As such, a MLTE-managed network offers a better than best-effort service. Many types of traditional and novel services are shifting toward IP/MPLS technology. Consequentially, MLTE algorithms and strategies should be conceived with the characteristics of such services in mind. We present a MLTE strategy that can be implemented in a robust and distributed way. This strategy is then taken as the starting point in a study which evaluates its suitability to such services. We show how the strategy can be adapted considering service performance metrics such as end-to-end delay, traffic loss, and routing stability, and how such service optimizations impact general MLTE objectives such as IP/MPLS logical topology mesh size reduction.

Multilayer traffic engineering for GMPLS-enabled networks

In recent years, significant work has been completed on traffic engineering enhancements to the generalized multiprotocol label switching protocol suite (E. Mannie Oct 2004) (D. Katz et al., Sept. 2003) (K. Kompella et al., Oct 2003). As a next step, reproducing the current trend of switching layers' integration happening in the data plane, network control is foreseen to go beyond the traditional per layer approach and tend toward an integrated model (K. Shimoto et al., Oct 2004) (E. Dotaro et al., Dec. 2004). In these multilayer environments, a single GMPLS control plane drives various distinct switching layers at the same time and as a coherent whole, taking benefit from the "common" property of GMPLS. Beyond this application of supporting network control across different technologies, in this article we catalog the unified traffic engineering paradigms, discuss their applicability, and present their enforcement techniques. Furthermore, we show that the common GMPLS concept has the advantage of low operational complexity, and enables unified TE capabilities such as efficient network resource usage and rapid service provisioning.     

GATEway: symbiotic inter-domain traffic engineering

There are a group of problems in networking that can most naturally be described as optimization problems (network design, traffic engineering, etc.). There has been a great deal of research devoted to solving these problems, but this research has been concentrated on intra-domain problems where one network operator has complete information and control. An emerging field is inter-domain engineering, for instance, traffic engineering between large autonomous networks. Extending intra-domain optimization techniques to inter-domain problems is often impossible without the information available within a domain, and providers are often unwilling to share such information. This paper presents an alternative: we propose a method for traffic engineering that does not require sharing of important information across domains. The method extends the idea of genetic algorithms to allow symbiotic evolution between two parties. Both parties may improve their performance without revealing their data, other than what would be easily observed in any case. We show the method provides large reductions in network congestion, close to the optimal shortest path routing across a pair of networks. The results are highly robust to measurement noise, the method is very flexible, and it can be applied using existing routing.     

IP network configuration for intradomain traffic engineering

The smooth operation of the Internet depends on the careful configuration of routers in thousands of autonomous systems throughout the world. Configuring routers is extremely complicated because of the diversity of network equipment, the large number of configuration options, and the interaction of configuration parameters across multiple routers. Network operators have limited tools to aid in configuring large backbone networks. Manual configuration of individual routers can introduce errors and inconsistencies with unforeseen consequences for the operational network. In this article we describe how to identify configuration mistakes by parsing and analyzing configuration data extracted from the various routers. We first present an overview of IP networking from the viewpoint of an Internet service provider and describe the kinds of errors that can appear within and across router configuration files. To narrow the scope of the problem, we then focus our attention on the configuration commands that relate to traffic engineering-tuning the intradomain routing protocol to control the flow of traffic through the ISP network. We present a case study of a prototype tool, developed in collaboration with AT&T IP Services, for checking the configuration of the AT&T IP Backbone and providing input to other systems visualization and traffic engineering     

NetScope: traffic engineering for IP networks

Managing large IP networks requires an understanding of the current traffic flows, routing policies, and network configuration. However, the state of the art for managing IP networks involves manual configuration of each IP router, and traffic engineering based on limited measurements. The networking industry is sorely lacking in software systems that a large Internet service provider can use to support traffic measurement and network modeling, the underpinnings of effective traffic engineering. This article describes the AT&T Labs NetScope, a unified set of software tools for managing the performance of IP backbone networks. The key idea behind NetScope is to generate global views of the network on the basis of configuration and usage data associated with the individual network elements. Having created an appropriate global view, we are able to infer and visualize the networkwide implications of local changes in traffic, configuration, and control. Using NetScope, a network provider can experiment with changes in network configuration in a simulated environment rather than the operational network. In addition, the tool provides a sound framework for additional modules for network optimization and performance debugging. We demonstrate the capabilities of the tool through an example traffic engineering exercise of locating a heavily loaded link, identifying which traffic demands flow on the link, and changing the configuration of intradomain routing to reduce the congestion     

MPLS and traffic engineering in IP networks

Rapid growth and increasing requirements for service quality, reliability, and efficiency have made traffic engineering an essential consideration in the design and operation of large public Internet backbone networks. Internet traffic engineering addresses the issue of performance optimization of operational networks. A paramount objective of Internet traffic engineering is to facilitate the transport of IP traffic through a given network in the most efficient, reliable, and expeditious manner possible. Historically, traffic engineering in the Internet has been hampered by the limited functional capabilities of conventional IP technologies. Recent developments in multiprotocol label switching (MPLS) and differentiated services have opened up new possibilities to address some of the limitations of the conventional technologies. This article discusses the applications of MPLS to traffic engineering in IP networks     

Decentralized optimal traffic engineering in connectionless networks

This work addresses the problem of optimal traffic engineering in a connectionless autonomous system. Based on nonlinear control theory, the approach taken in This work provides a family of optimal adaptation laws. These laws enable each node in the network to independently distribute traffic among any given set of next hops in an optimal way, as measured by a given global utility function of a general form. This optimal traffic distribution is achieved with minimum information exchange between neighboring nodes. Furthermore, this approach not only allows for optimal multiple forwarding paths but also enables multiple classes of service, e.g., classes of service defined in the differentiated services architecture. Moreover, the proposed decentralized control scheme enables optimal traffic redistribution in the case of link failures. Suboptimal control laws are also presented in an effort to reduce the computational burden imposed on the nodes of the network. Finally, an implementation of these laws with currently available technology is discussed.     

ATM traffic engineering for ABR service provisioning

The Available Bit Rate (ABR) service is being designed as a low-cost transport service over ATM, which will be using the bandwidth left available after servicing connections of another, high-priority class. The implementation of the ABR service requires large buffers at each multiplexing/switching stage to keep cell-loss rates down to a minimum, and a feedback mechanism from the network to the terminals in order for the latter to adjust their traffic profiles according to the prevailing congestion conditions. Thus, an enhanced set of traffic control functions is necessary to support this new service. In this paper the main traffic analysis and control problems related with the ABR service are addressed, modelled and answered on the basis of effective rates defined for the multiplexed connections. Emphasis is given to a simple CAC scheme which consists in allocating peak rates to the high-priority class and effective rates to the ABR class. An adaptive shaping mechanism is then required to enforce the contracted effective rates for the ABR streams. Producing ON/OFF streams facilitates the control functions by allowing the use of approximate closedform calculations.This work has been carried out partly in the framework of the RACE EXPLOIT project.     

On route selection for interdomain traffic engineering

In this article we investigate a model of route selection for interdomain traffic engineering where routing to multiple destinations can be coordinated. We identify potential routing instability and inefficiency problems, and derive a set of practical guidelines to guarantee stability without global coordination. Using a realistic Internet topology, we show that route oscillations can happen even when a small number of ASes coordinate route selection for just a small number of destinations if the coordination does not follow our guidelines. Wc further extend our model so that ASes can adopt any route selection algorithms in a class of algorithms we call rational route selection algorithms; and the local ranking of routes of an AS can depend on ingress traffic patterns. We show that persistent route oscillations can happen in certain network settings even if the ASes strictly follow the constraints imposed by business considerations, and adopt any rational route selection algorithms.