Interdomain traffic engineering with BGP

Traffic engineering is performed by means of a set of techniques that can be used to better control the flow of packets inside an IP network. We discuss the utilization of these techniques across interdomain boundaries in the global Internet. We first analyze the characteristics of interdomain traffic on the basis of measurements from three different Internet service providers and show that a small number of sources are responsible for a large fraction of the traffic. Across interdomain boundaries, traffic engineering relies on a careful tuning of the route advertisements sent via the border gateway protocol. We explain how this tuning can be used to control the flow of incoming and outgoing traffic, and identify its limitations.     

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.     

A distributed mobility control scheme in LISP networks

The locator identifier separation protocol (LISP) has been made as an identifier-locator separation scheme for scalable Internet routing. However, the LISP was originally designed for fixed network environment, rather than for mobile network environment. In particular, the existing LISP mobility control schemes use a centralized map server to process all the control traffics, and thus they are intrinsically subject to some limitations in mobile environment, such as large overhead of mapping control traffics at central map server and degradation of handover performance. To overcome these problems, we propose a distributed mobility control scheme in LISP networks. In the proposed scheme, we assume that a mobile host has a hierarchical endpoint identifier which contains the information of its home network domain. Each domain has a distributed map server (DMS) for distributed mapping management of Endpoint Identifiers (EIDs) and Locators (LOCs). For roaming support, each DMS maintains a home EID register and a visiting EID register which are used to keep the EID-LOC mappings for mobile hosts in the distributed manner. For performance analysis, we compare the control traffic overhead (CTO) at map servers, the signaling delay required for EID-LOC mapping management, and the handover delay for the existing and proposed schemes. From numerical results, it is shown that the proposed distributed scheme can give better performance than the existing centralized schemes in terms of CTO, total signaling delay for EID-LOC mapping management, and handover delay.     

Delayed Internet routing convergence

This paper examines the latency in Internet path failure, failover, and repair due to the convergence properties of interdomain routing. Unlike circuit-switched paths which exhibit failover on the order of milliseconds, our experimental measurements show that interdomain routers in the packet-switched Internet may take tens of minutes to reach a consistent view of the network topology after a fault. These delays stem from temporary routing table fluctuations formed during the operation of the border gateway protocol (BGP) path selection process on the Internet backbone routers. During these periods of delayed convergence, we show that end-to-end Internet paths will experience intermittent loss of connectivity, as well as increased packet loss and latency. We present a two-year study of Internet routing convergence through the experimental instrumentation of key portions of the Internet infrastructure, including both passive data collection and fault-injection machines at major Internet exchange points. Based on data from the injection and measurement of several hundred thousand interdomain routing faults, we describe several unexpected properties of convergence and show that the measured upper bound on Internet interdomain routing convergence delay is an order of magnitude slower than previously thought. Our analysis also shows that the upper theoretic computational bound on the number of router states and control messages exchanged during the process of BGP convergence is factorial with respect to the number of autonomous systems in the Internet. Finally, we demonstrate that much of the observed convergence delay stems from specific router vendor implementation decisions and ambiguity in the BGP specification     

Implications of Autonomy for the Expressiveness of Policy Routing

Thousands of competing autonomous systems must cooperate with each other to provide global Internet connectivity. Each autonomous system (AS) encodes various economic, business, and performance decisions in its routing policy. The current interdomain routing system enables each AS to express policy using rankings that determine how each router in the AS chooses among different routes to a destination, and filters that determine which routes are hidden from each neighboring AS. Because the Internet is composed of many independent, competing networks, the interdomain routing system should provide autonomy, allowing network operators to set their rankings independently, and to have no constraints on allowed filters. This paper studies routing protocol stability under these conditions. We first demonstrate that ldquonext-hop rankings,rdquo commonly used in practice, may not ensure routing stability. We then prove that, when providers can set rankings and filters autonomously, guaranteeing that the routing system will converge to a stable path assignment imposes strong restrictions on the rankings ASes are allowed to choose. We discuss the implications of these results for the future of interdomain routing.     

LISP controller: a centralized LISP management system for ISP networks

As the current Internet architecture is suffering from scalability issues, the network research community has proposed alternative designs for the Internet architecture. Among those solutions that adopt the idea of locator/identifier split paradigm, the locator/identifier separation protocol (LISP) has been considered as the most promising solution because of its incrementally deployable feature. Despite various advantages provided by LISP, many ISPs are still conservative to adopt LISP into their production network because the standard LISP does not fully satisfy ISP's requirements on LISP‐enabled services. In this paper, we define ISP's requirements on LISP‐enabled commercial services and describe limitations of the standard LISP from an ISP's perspective. Also, we propose LISP controller, a centralized LISP management system. By using LISP controller, we evaluate three ISP's representative LISP use cases: traffic engineering, virtual machine live migration, and vertical handover. The results show that the proposed LISP controller provides centralized management, controllability, and fast map entry update, without any modifications on the standard LISP. LISP controller allows an ISP to control and manage its LISP‐enabled services while satisfying ISP's requirements. Copyright © 2015 John Wiley & Sons, Ltd.     

An overview of routing optimization for internet traffic engineering

Traffic engineering is an important mechanism for Internet network providers seeking to optimize network performance and traffic delivery. Routing optimization plays a key role in traffic engineering, finding efficient routes so as to achieve the desired network performance. In this survey we review Internet traffic engineering from the perspective of routing optimization. A taxonomy of routing algorithms in the literature is provided, dating from the advent of the TE concept in the late 1990s. We classify the algorithms into multiple dimensions: unicast/multicast, intra-/inter- domain, IP-/MPLS-based and offline/online TE schemes. In addition, we investigate some important traffic engineering issues, including robustness, TE interactions, and interoperability with overlay selfish routing. In addition to a review of existing solutions, we also point out some challenges in TE operation and important issues that are worthy of investigation in future research activities.     

SDI: a multi-domain SDN mechanism for fine-grained inter-domain routing

Software-defined networking (SDN) scheme decouples network control plane and data plane, which can improve the flexibility of traffic management in networks. OpenFlow is a promising implementation instance of SDN scheme and has been applied to enterprise networks and data center networks in practice. However, it has less effort to spread SDN control scheme over the Internet to conquer the ossification of inter-domain routing. In this paper, we further innovate to the SDN inter-domain routing inspired by the OpenFlow protocol. We apply SDN flow-based routing control to inter-domain routing and propose a fine-granularity inter-domain routing mechanism, named SDI (Software Defined Inter-domain routing). It enables inter-domain routing to support the flexible routing policy by matching multiple fields of IP packet header. We also propose a method to reduce redundant flow entries for inter-domain settings. And, we implement a prototype and deploy it on a multi-domain testbed.     

Performance Analysis of Infrastructure Service Provision with GMPLS-Based Traffic Engineering

Dynamic sharing of the common physical network is envisioned as a key enabler for the emerging Internet technologies. This paper addresses challenges related to resource sharing in the physical layer and analyzes the performance of infrastructure service provision with control plane mechanisms based on generalized multi protocol label switching (GMPLS). In our approach, the provisioning of infrastructure services is supported by two novel concepts for GMPLS traffic engineering (TE): resource visibility and inter-domain exchange. Resource visibility is a new network control plane concept, which defines the usage polices for transmission, multiplexing, and switching resources in multiple GMPLS layers. In our architecture, every network resource may exhibit different visibility to different services at different layers. The inter-domain exchange, here referred to as GMPLS exchange point (GXP), is the physical layer equivalent of the Internet exchange point (IXP). Just as how the IXP manages interconnections of autonomous systems (AS) in the Internet, the GXP manages dynamic interconnections of multiple provider domains and enables them to advertise their physical resources to other domains. We model the dynamic provisioning of infrastructure services using graph theory and deploy GMPLS traffic engineering (TE) to optimize the routing and resource yields. The results obtained demonstrate that traffic engineering with resource visibility and GXP brings significant performance benefits in resource utilization and infrastructure extensibility, especially when network providers set up LSPs as a result of collaborative and carrier-neutral traffic engineering where they share information about resource capabilities and utilization     

A policy based framework for quality of service management in software defined networks

Growth in multimedia traffic over the Internet increases congestion in the network architecture. Software-Defined Networking (SDN) is a novel paradigm that solves the congestion problem and allows the network to be dynamic, intelligent, and it centrally controls the network devices. SDN has many advantages in comparison to traditional networks, such as separation of forwarding and control plane from devices, global centralized control, management of network traffic. We design a policy-based framework to enhance the Quality of Service (QoS) of multimedia traffic flows in a potential SDN environment. We phrase a max-flow-min-cost routing problem to determine the routing paths and presented a heuristic method to route the traffic flows in the network in polynomial time. The framework monitors the QoS parameters of traffic flows and identifies policy violations due to link congestion in the network. The introduced approach dynamically implements policy rules to SDN switches upon detection of policy violations and reroutes the traffic flows. The results illustrate that the framework achieves a reduction in end-to-end delay, average jitter, and QoS violated flows by 24%, 37%, and 25%, respectively, as compared to the Delay Minimization method. Furthermore, the proposed approach has achieved better results when compared to SDN without policy-based framework and reduced end-to-end delay, average jitter, and QoS violated flows by 51%, 62%, and 28%, respectively.

     

Prediction-based routing as RWA in multilayer traffic engineering

Multilayer traffic engineering (MLTE) allows coping with ever-increasing and varying traffic demands in IP-over-Optical multilayer networks. It utilizes cross-layer TE (Traffic Engineering) techniques to provision optical lightpath capacity to the IP/MPLS (Internet Protocol/ Multi-Protocol Label Switching) logical topology on-demand. Such provisioning however causes optical connection arrival rates that pose strong performance requirements to Routing and Wavelength Assignment (RWA) strategies. Collecting up-to-date network information for the RWA with rapidly changing network states can be quite difficult. Exposing optical layer state information to the IP layer in the overlay model, or transforming this optical layer information in a workable representation in an integrated control plane is similarly problematic. Prediction-Based Routing (PBR) has been proposed as a RWA mechanism for optical transport networks; it bases routing not on possibly inaccurate or outdated network state, but instead on previous connections set-up. In this article, we propose to implement PBR as the RWA mechanism in the optical layer of a multilayer network, and use the predictive capabilities of PBR to expose dynamic optical network information into the multilayer traffic engineering algorithm with minimal control plane overhead. Some simulations show the benefits of using the PBR in the optical layer for MLTE purposes.     

Deriving traffic demands for operational IP networks: methodologyand experience

Engineering a large IP backbone network without an accurate network-wide view of the traffic demands is challenging. Shifts in user behavior, changes in routing policies, and failures of network elements can result in significant (and sudden) fluctuations in load. We present a model of traffic demands to support traffic engineering and performance debugging of large Internet service provider networks. By defining a traffic demand as a volume of load originating from an ingress link and destined to a set of egress links, we can capture and predict how routing affects the traffic traveling between domains. To infer the traffic demands, we propose a measurement methodology that combines flow-level measurements collected at all ingress links with reachability information about all egress links. We discuss how to cope with situations where practical considerations limit the amount and quality of the necessary data. Specifically, we show how to infer interdomain traffic demands using measurements collected at a smaller number of edge links-the peering links connecting to neighboring providers. We report on our experiences in deriving the traffic demands in the AT&T IP Backbone, by collecting, validating, and joining very large and diverse sets of usage, configuration, and routing data over extended periods of time. The paper concludes with a preliminary analysis of the observed dynamics of the traffic demands and a discussion of the practical implications for traffic engineering     

Performance analysis of a traffic engineering solution for multilayer networks based on the GMPLS paradigm

This paper presents an integrated traffic engineering (TE) system for new generation multilayer networks based on the generalized multiprotocol label switching (GMPLS) paradigm, and reports the performance analysis of such a system. The proposed TE system aims at dynamically reacting to traffic changes and, at the same time, fulfilling quality of service (QoS) requirements for different classes of service. The proposed solution consists of a hybrid routing approach and a bandwidth management strategy. The former makes use of both off-line and on-line methods to accommodate traffic requests. The latter, based on an "elastic" use of the bandwidth, allows the handling of different priorities among data flows, possible preemptions, and rerouting. The proposed TE permits the accommodation of the largest amount of traffic, while guaranteeing good performance to mission-critical services. The main building blocks and the operations of the system are reported and the major advantages are discussed. The performance of the proposed system are compared with the ones relating to a reasonable alternative system based on overprovisioning, to highlight its advantages of in terms of traffic volume that can be accommodated for a given network infrastructure.     

Performance and testbed study of topology reconfiguration in IP over optical networks

With the widespread deployment of Internet protocol/wavelength division multiplexing (IP/WDM) networks, it becomes necessary to develop traffic engineering (TE) solutions that can effectively exploit WDM reconfigurability. More importantly, experimental work on reconfiguring lightpath topology over testbed IP/WDM networks is needed urgently to push the technology forward to operational networks. This paper presents a performance and testbed study of topology reconfiguration for IP/WDM networks. IP/WDM TE can be fulfilled in two fashions, overlay vs. integrated, which drives the network control software, e.g., routing and signaling protocols, and selects the corresponding network architecture model, e.g., overlay or peer-to-peer. We present a traffic management framework for IP over reconfigurable WDM networks. Three "one-hop traffic maximization"-oriented heuristic algorithms for lightpath topology design are introduced. A reconfiguration migration algorithm to minimize network impact is presented. To verify the performance of the topology design algorithms, we have conducted extensive simulation study. The simulation results show that the topologies designed by the reconfiguration algorithms outperform the fixed topology with throughput gain as well as average hop-distance reduction. We describe the testbed network and software architecture developed in the Defense Advanced Research Projects Agency (DARPA) Next Generation Internet (NGI) SuperNet Network Control and Management project and report the TE experiments conducted over the testbed.     

基于SDN的互联网域间路由研究

互联网流量的爆发式增长,叠加互联网流量固有的突发性特点,使得网络流量不均衡现象日益加剧。传统BGP协议由于缺乏全网拓扑和全局流量观,只能遵循标准BGP选路原则,在解决流量调度和负载均衡方面存在不足。针对BGP协议存在的局限性,研发了基于RR+的互联网骨干网流量调度系统,并应用于ChinaNet骨干网的网内中继、网间互联出口、IDC出口等多个流量优化场景。更进一步地,提出了一种基于SDN的互联网域间路由架构,通过在域间控制器之间交换BGP路由,无需在域内和域间运行BGP协议,极大地简化了网络协议,并能够实现灵活的流量调度和负载均衡。     

Distributed schemes for diverse path computation in multidomain MPLS networks

MPLS is currently used by several JSPs to carry some high-value traffic components, such as telephony over IP trunks and VPNs. For this type of traffic, service availability is a critical QoS dimension that needs to be protected from network failures. With MPLS-TE, this can be achieved by means of path protection schemes, where active and backup LSPs are routed along diverse paths. Besides protection, path diversity can be exploited for load balancing, another common means of QoS improvement. In order to preserve other QoS requirements, the paths must meet certain constraints (e.g., bandwidth availability, low load) and/or minimize some metric (e.g., hop count). This requires the ability to establish path diversity in an optimal way. In many cases of practical interest, the QoS traffic has an interdomain scope. This is the case for ToIP and VPN traffic between different carriers, or between different ASs owned by the same carrier, as found, for example, after corporate acquisitions or mergers. Therefore, path diversity is a requirement for interdomain traffic engineering. In this work we address path diversity in a multidomain network, where individual domains are capable of connection-oriented forwarding and endowed with an MPLS-TE control plane. For administrative and/or scalability reasons intradomain routing information is not disseminated externally, so dynamic path computation must be achieved by a distributed scheme based on interdomain collaboration. We briefly describe three alternative schemes recently proposed for interdomain diverse path computation, and quantitatively assess their performance with simulations over real ISP topologies.     

A Survey of BGP Security Issues and Solutions

As the Internet's de facto interdomain routing protocol, the Border Gateway Protocol (BGP) is the glue that holds the disparate parts of the Internet together. A major limitation of BGP is its failure to adequately address security. Recent high-profile outages and security analyses clearly indicate that the Internet routing infrastructure is highly vulnerable. Moreover, the design of BGP and the ubiquity of its deployment have frustrated past efforts at securing interdomain routing. This paper considers the current vulnerabilities of the interdomain routing system and surveys both research and standardization efforts relating to BGP security. We explore the limitations and advantages of proposed security extensions to BGP, and explain why no solution has yet struck an adequate balance between comprehensive security and deployment cost.     

A Multi-link Mechanism for Heterogeneous Radio Networks

Nowadays, smart mobile devices drive the mobile traffic growth rapidly. Most smart mobile devices are equipped with multiple radio network interfaces, such as High Speed Packet Access (HSPA), Long Term Evolution (LTE), and Wi-Fi. Therefore, integration of multiple networks is a viable solution to fulfill traffic offloading and the Quality-of-Service (QoS) requirement of data usage for mobile users. In this paper, we propose a multi-link mechanism to handle the radio network selection and switching between LTE and Wi-Fi networks. A Multi-Link Adaptor (MLA) and a Multi-Connection Manager (MCM) are proposed for the User Equipment (UE) and the core network, respectively, to handle the multi-link mechanism. The applications executed in the UEs do not need to be modified under the proposed approach. The MLA maintains a QoS class table and a routing table for the network selection procedure and uses the GPRS Tunneling Protocol-Control plane (GTP-C) control messages to execute network switching. In the future, we will measure the throughput of the multi-link network and the switch delay between the heterogeneous radio networks.     

LISP-MN: Mobile Networking Through LISP

The current Internet architecture was not designed to easily accommodate mobility because IP addresses are used both to identify and locate hosts. The Locator/Identifier Separation Protocol (LISP) decouples them by considering two types of addresses: EIDs that identify hosts, and RLOCs that identify network attachment points and are used as routing locators. LISP, with such separation in place, can also offer native mobility. LISP-MN is a particular case of LISP which specifies mobility. In this paper we provide a comprehensive tutorial on LISP-MN, showing its main features and how it compares to existing mobility protocols.