Internet QoS

Architectures and Mechanisms for Quality of Service


  • Zheng Wang

Guaranteeing performance and prioritizing data across the Internet may seem nearly impossible because of an increasing number of variables that can affect and undermine service. But if you're involved in developing and implementing streaming video or voice, or other time-sensitive Internet applications, you understand exactly what's at stake in establishing Quality of Service (QoS) and recognize the benefits it will bring to your company.

What you need is a reliable guide to the latest QoS techniques that addresses the Internet's special challenges. Internet QoS is it-the first book to dig deep into the issues that affect your ability to provide performance and prioritization guarantees to your customers and users! This book gives a comprehensive view of key technologies and discusses various analytical techniques to help you get the most out of network resources as you strive to make, and adhere to, meaningful QoS guarantees.

View full description


network engineers, designers, and managers, equipment vendors, service providers and ISPs


Book information

  • Published: March 2001
  • ISBN: 978-1-55860-608-1


"What I really like about this book is that it cuts through the vast amount of noise about QoS in the Internet, and pulls out the core ideas (integrated and differentiated services, MPLS, and traffic engineering) in plain and simple technical prose. Internet QoS presents a balanced view of the various technologies and relates them to their practical use. It's up to date, but unlikely to go out of date quickly either, so should prove useful to engineers and students alike."
—Jon Crowcroft, University College London

Table of Contents

PREFACE1 THE BIG PICTURE1.1 RESOURCE ALLOCATION1.1.1 Integrated Services1.1.2 Differentiated Services1.2 PERFORMANCE OPTIMIZATION1.2.1 Multiprotocol Label Switching1.2.2 Traffic Engineering1.3 SUMMARYFURTHER READING2 INTEGRATED SERVICES2.1 INTRODUCTION2.2 REAL-TIME APPLICATIONS2.2.1 Playback Applications2.2.2 Tolerant and Intolerant Playback Applications2.2.3 Lessons fromMBONE2.3 INTEGRATED SERVICES ARCHITECTURE2.3.1 Basic Approach2.3.2 Key Components2.4 SERVICE MODELS2.4.1 Flow Specification2.4.2 Guaranteed Service2.4.3 Controlled Load Service2.5 RESOURCE RESERVATION SETUP (RSVP)2.5.1 Basic Features2.5.2 Operation Overview2.5.3 RSVP Messages2.5.4 Reservation Styles 2.5.5 Adspec, OPWA, and Slack Term2.6 FLOW IDENTIFICATION2.6.1 Basic Requirements2.6.2 Design Choices2.6.3 Hashing-Based Schemes2.6.4 Performance Evaluation2.7 PACKET SCHEDULING2.7.1 Basic Requirements2.7.2 Design Choices2.7.3 Weighted Fair Queuing2.7.4 Variants of WFQ2.8 INTEGRATED SERVICES OVER SPECIFIC LINK LAYERS2.8.1 Local Area Networks2.8.2 ATM Networks2.9 SUMMARYFURTHER READING3 DIFFERENTIATED SERVICES3.1 INTRODUCTION3.2 DIFFERENTIATED SERVICES FRAMEWORK3.2.1 Basic Approach3.2.2 Service and Forwarding Treatment3.2.3 Per-Hop Behaviors (PHBs)3.2.4 Services3.3 DIFFERENTIATED SERVICES FIELD3.3.1 Structure of DS Field3.3.2 Historical Codepoint Definition3.3.3 Current Codepoint Allocation3.4 TRAFFIC CLASSIFICATION AND CONDITIONING3.4.1 Classifier3.4.2 Traffic Conditioner3.4.3 Location of Traffic Classifiers and Conditioners3.4.4 Configuring Traffic Classifiers and Conditioners3.5 ASSURED FORWARDING3.5.1 AF PHB Group3.5.2 Implementation Guideline3.5.3 Example Services3.6 EXPEDITED FORWARDING3.6.1 EF PHB3.6.2 Implementation Guideline3.7 INTEROPERABILITY WITH NON-DS-COMPLIANT NETWORKS3.7.1 Non-DS-Compliant Node within a DS Domain3.7.2 Transit Non-DS-Capable Domain3.8 PACKET CLASSIFICATION3.8.1 Basic Requirements3.8.2 Classification Algorithms3.9 TRAFFIC POLICING3.9.1 Metering and Marking3.9.2 Dual Token Bucket Algorithm3.10 END-TO-END RESOURCE MANAGEMENT3.10.1 Integrated Services over Differentiated Services3.10.2 Interdomain Bandwidth Allocation3.10.3 End-System Congestion Control 3.11 PERFORMANCE ISSUES IN DIFFERENTIATED SERVICES3.11.1 Network Configuration3.11.2 Traffic Profiles and Bottleneck Bandwidth 3.11.3 Protection from Misbehaving Sources3.12 SUMMARYFURTHER READING4 MULTIPROTOCOL LABEL SWITCHING4.1 INTRODUCTION4.2 MOTIVATION4.2.1 IP over ATM Integration4.2.2 Simpler Forwarding Paradigm4.2.3 Traffic Engineering4.3 OVERVIEW 4.3.1 Routing vs. Switching4.3.2 Label-Switching Proposals 4.3.3 Comparison of Approaches4.4 MPLS ARCHITECTURE 4.4.1 Key Concepts4.4.2 Forwarding Equivalency Classes 4.4.3 Hierarchy and Label Stacking 4.4.4 Label Stack Encoding 4.4.5 Loop Detection4.5 LABEL DISTRIBUTION PROTOCOLS 4.5.1 LDP4.5.2 CR-LD4.5.3 RSVP-TE4.5.4 Comparison4.6 SUMMARYFURTHER READING5 INTERNETTRAFFIC ENGINEERING5.1 INTRODUCTION5.2 THE FISH PROBLEM 5.3 TRAFFIC-ENGINEERING SOLUTIONS5.4 OPTIMIZATION OBJECTIVES 5.5 BUILDING BLOCKS 5.5.1 Data Repository5.5.2 Topology and State Discovery5.5.3 Traffic Demand Estimation5.5.4 Route Computation 5.5.5 Network Interface5.6 TOPOLOGY AND STATE DISCOVERY 5.7 CONSTRAINT-BASED ROUTING 5.7.1 Mathematical Formulation5.7.2 Overlay Model 5.7.3 Peer Model5.8 MULTIPATH LOAD SHARING 5.8.1 Direct Hashing5.8.2 Table-Based Hashing5.9 SUMMARY FURTHER READINGCLOSING REMARKS GLOSSARY BIBLIOGRAPHY INDEX ABOUT THE AUTHOR