Foreword
Congestion control is unquestionably one of the most important, most fundamental topics in computer networking. It’s also one of most challenging, as it requires controlling endpoints that are potentially distributed around the globe, in different organizations, and supporting different applications. The role of the network layer in supporting transport-layer congestion control is also a multi-faceted, nuanced challenge. And congestion control is needed in just about every Internet scenario one can imagine: from the public Internet that spans the globe and carries all types of traffic, to long “fat” pipes carrying massive amounts of file-transfer data, to specialized datacenter networks, to private commercial backbone networks, to mobile and wireless networks.
With all of these challenges, how does one make sense of the many (many!) approaches towards congestion control that have been developed? What are the fundamental challenges these approaches are solving? What is the role of the network layer, and more broadly what is the design space for congestion control protocols? Are there broad classes or approaches towards congestion control that can be identified? Which approaches have been adopted in practice, and why? And among those many “flavors”/variations of TCP that you might have heard about—how do they differ and in what scenarios are they best used, and why? So many questions!
To make sense of this and to answer all of these questions (and more) would require not just a book, but a great book! And now fortunately, there is such a book—this book! The three authors of TCP Congestion Control: A Systems Approach are among the most knowledgeable congestion control researchers on the planet—Brakmo and Peterson’s TCP Vegas protocol (you can learn more about that in section 5.1) pioneered the notion that endpoints could anticipate and avoid congestion, rather than react to observed congestion; TCP Vegas has served as a foundation on which more recent congestion avoidance protocols (such as the BBR protocol championed by Google, see Section 5.3) have been designed. The authors are also absolutely fabulous writers (and I say this as a textbook author myself)—lucid, clear, and engaging, and able to organize and communicate complex ideas, with just the right amount of detail and discussion of practice. The “systems approach” that Larry Peterson and Bruce Davie have championed is also exactly what is needed to truly understand congestion control, where deep, system-wide issues in network architecture come to the fore (e.g., the separation and interaction of network and transport layer functionalities; the question of implementing networks services, such as congestion control, in the either the application layer or in the network).
This book is a needed and most welcome addition to the fabulous set of open source, “systems approach” books that Larry, Bruce and others have been developing. I hope you read it cover-to-cover, consult it again later as you need it in the future, and enjoy it as much as I have.