Network Coding for Engineers (eBook)

Muriel Médard, Vipindev Adat Vasudevan, Morten Videbæk Pedersen, Ken R. Duffy (Autoren)

eBook Download: EPUB

2025
411 Seiten
Wiley-IEEE Press (Verlag)
978-1-394-21729-8 (ISBN)

Infos

Understand the fundamentals of network coding from an engineering perspective with this accessible guide

Network Coding is a method of increasing network throughput and efficiency by encoding and decoding transmitted data packets instead of simply forwarding them. It was mainly a body of information theory until the rise of random linear networking coding (RLNC), a method ideally suited to wireless networks and other cooperative environments. The ease of introducing network coding to legacy systems and the resulting gains in efficiency have made this a widely applied technology with the potential to revolutionize networked communications.

Network Coding for Engineers introduces the fundamentals of this exciting subject from an engineering perspective. Beginning with the basics, including step-by-step details for implementing network coding and current applications, it also highlights potential uses of network coding in the communications technologies of the future. The result is an innovative and accessible introduction to a subject quickly becoming indispensable.

Network Coding for Engineers readers will also find:

• A structure that facilitates gradual deepening of knowledge, ideal for students and new readers
• Follows a semester-long course curriculum structure, making it suitable for direct adaptation for academic purposes
• Detailed discussion of future applications in technology areas including post-quantum cryptography, 6G, and more
• Design principles for different network models, such as multi-path and mesh networks

Network Coding for Engineers is ideal for electrical engineering and computer science students, particularly those studying advanced networking and communications and related subjects.

Muriel Médard, PhD, is NEC Professor of Software Science and Engineering in the Electrical Engineering and Computer Science (EECS) Department and Head of the Network Coding and Reliable Communications Group in the Research Laboratory of Electronics at the Massachusetts Institute of Technology (MIT), USA. She is a member of the US National Academy of Engineering and a Fellow of the US National Academy of Inventors and the Institute of Electrical and Electronics Engineers, among other organizations, and is a past president of the IEEE Information Theory Society.

Vipindev Adat Vasudevan, PhD, is a Postdoctoral Associate in the Network Coding and Reliable Communications Group in the Research Laboratory of Electronics, Massachusetts Institute of Technology (MIT), USA. He has previously worked as an MSCA Fellow in the Wireless Telecommunications Laboratory of the University of Patras, Greece, and has researched extensively into network coding, network security, 5G, and related subjects.

Morten Videbæk Pedersen, PhD, is co-founder and Chief Technology Officer of Steinwurf APS, a company specializing in industrial-grade high-performance content distribution systems, where he focuses on low-complexity network coding algorithms and cooperative networking programs. He is a member of the Mobile Devices Research Group, Aalborg, Denmark, and a past recipient of the Forum Nokia Developer Champion Award.

Ken R. Duffy, PhD, is a Professor of Electrical and Computer Engineering, and a Professor of Mathematics, at Northeastern University (NU), USA. He leads the Engineering Probability Information and Communications Laboratory and is core faculty at NU's Institute for the Wireless Internet of Things. He was previously the Director of the Hamilton Institute at the National University of Ireland, Maynooth, and a co-director of the Science Foundation Ireland Centre for Research Training in Foundations of Data Science.

Textauszug

1
Introduction

The objective of packet networks, such as the Internet, is to reliably transport information from sources to receivers. While data is packaged for communication, which may involve compressing it and adding headers describing destinations as well as other information, traditionally the data streams themselves are treated as immutable as they traverse the network, in the sense that the information of individual streams is kept separate and intact throughout their transit. The simple premise of Network Coding (NC) is that data can be readily algebraically manipulated and that making use of that fact results in a relaxation of many networking problems, which can be leveraged to vastly improve performance in a number of distinct dimensions.

As a basic thought experiment, if the network wishes to transport two pieces of information, and , it could instead transport and and allow the receiver to solve a basic set of linear equations, i.e. , to recover both. Why would that be useful? Think of a setting where three packets are transmitted across a network, but one will get lost in transit, as can happen in networks such as the Internet. If the network transmits , , and , no matter which two packets get through, the receiver can reconstruct the original data. The generalization and exploitation of this simple principle results in large, practically achievable gains in performance, as we shallsee.

NC remained in the realm of pure information theory until the advent of Random Linear Network Coding (RLNC) [1]. It has evolved from simple modulo additions of two packets to creating linear combinations of multiple packets in a finite field and communicating the digital result as doing so enables significant improvement in the bandwidth efficiency of networks. RLNC inherently generates robustness and adaptability in dynamic environments [2]. RLNC has proven suitable for distributed, dynamic environments such as wireless networks. Future networks, such as small cell environments featuring Device‐to‐Device (D2D) communication and cooperation between devices, will have to ensure that every user in thenetwork is fairly provided with services, where distinct metrics concerning the quality of service, such as throughput, delay, and latency, all have to be met. In cooperative environments, RLNC can achieve the upper bound of efficiency in multicasting. RLNC has been established to be a practical solution in these settings that can provide higher throughput and reliability with lower latency over unreliable network infrastructure. A key feature of RLNC is that it can be readily integrated into legacy systems, thanks to its compatibility with existing protocols and its ability to be implemented both in software as well as hardware. While RLNC seeks to achieve optimum efficiency in terms of bandwidth usage by sending coded packets over different channels, it also naturally provides erasure correction and imparts resistance to man‐in‐the‐middle attacks. While the use of NC by itself provides only weak protections security [3], it has been extensively studied how, with some augmentation, it is possible to harness more benefits such as increased security and privacy protection using RLNC in the challenging, quantum computing times ahead.

In a world with increasing demands on multiple fronts, it becomes a necessity to re‐envision communication protocols and RLNC has already been proven to be a way ahead. Different industrial adopters are already benefiting from coding principles and use RLNC in a variety of applications. It has found its way to standards and industrial deployments [4, 5] in line with the significant theoretical research that has been happening around the topic in the last couple of decades. There have been great resources, not only the large number of papers published in reputed conferences and journals but also great books introducing different dimensions of RLNC in the literature. However, a textbook that assists the adaptation of NC from its excellent results in the literature to engineering solutions has not materialized. This book tries to bridge the gap between the academic advancements in the area of NC to its practical realizations, from a complete engineering perspective.

1.1 Vision and Outline

This book walks the reader through the nuances of practical NC and its implementation in real‐world applications with curated, directly relevant mathematical explanations. Excellent resources for theoretical explanation of the concepts are listed, but mainly as additional readings, as the book itself is self‐contained in the material it presents. It concentrates on how these concepts can be used to formulate engineering solutions in the complex communication scenarios that are expected as part of current and future networks. This book begins with the basics of NC and explores more advanced concepts step‐by‐step with implementation details and possible use cases. Further, it provides a look ahead on how NC can be used in varied scenarios such as post‐quantum cryptography and heterogeneous wireless networks. This textbook is written in a way that it can be a stand‐alone read for engineers or go along with a lab‐based semester or quarter course on NC. It expects minimal prior knowledge of communication and information theory and presents the concepts from the introductory level.

As a structured and gradual learning experience, the book lays out the concepts of NC from basics to its more complex adaptations in a simple but rigorous manner. This textbook is designed to suit the requirements of the course material for a lab‐based semester‐long course or a one‐quarter course, while it also allows communication engineers to use it for self‐study and explore RLNC‐based implementations. For a one‐quarter (9‐week) course, Chapters 1, 2, 4, and 5 will form the basis of a course with sufficient substance and detail to serve as a stand‐alone introduction. For such a course, it would be recommended to skip some of the optional material (marked with **) in Chapters 2 and 4. Portions marked with ** indicate that the material in them provides interesting theoretical background, but they can be omitted. The sections with ** will be of interest to students who are curious about some of the mathematical underpinnings of NC, but they are not required for implementing RLNC algorithms or for developing an understanding of how they can benefit the operation.

For a full one‐term (12‐week) course, a more software engineering slanted curriculum will incorporate the quarter course material discussed above with all of Chapter 2, and, additionally, all of Chapter 3, most of Chapter 5, and all of Chapter 6. A one‐term course oriented more towards network optimization and architecture will benefit from covering all of Chapter 2, as well as all of Chapter 4 and additionally Chapters 5 and 6. A one‐term course oriented towards a more theoretical audience will benefit from covering all of Chapter 2, as well as all of Chapters 4, 7, and 8.

The clear mapping between the book and class organization allows instructors or engineers following a self‐paced program, a clear and ready way to use the book. While the primary focus is on the application of NC, different network models such as multipath and mesh networks as well as their design principlesare covered.

The book can be used in modular way to meet the needs of different engineering goals. Engineers who intend to understand how to construct a detailed, low‐level coding language version of NC modules may benefit from reading portions marked with a *. These provide extensive algorithmic development of implementing finite field operations in a context that is useful to NC. Those sections can be skipped by readers whose goal is to understand NC and incorporate it into systems, but who intend to use commercially available packages, say KODO library, to manage the mechanics of NC.

The organization of the book is as follows. This first chapter focuses on introducing NC as the use equations that result from combining data rather than the original data in networks. The data allows flexibility which is not only convenient, it also is the root of the robustness, design flexibility, and theoretical optimality of NC in many settings. The chapter also introduces the reader to a Python‐based package that will readily manage erasures. This package will provide readers with a tool that will suffice to verify the majority of the engineering uses ofNC.

Chapter 2 introduces the reader to finite representations. The approach is rigorous, but entirely practical in its philosophy. The finite field results focus on explaining the arithmetic that is critical to NC implementations. This discussion will make sure the engineers get a clear picture of NC implementations and should be able to choose different designs according to their application scenario.

Chapter 3 provides the reader with a detailed guide to implementing NC. Together, Chapters 2 and 3 provide a detailed description of the finite field representations of the data required for the inner mechanics of NC. The chapter also includes detailed pseudo‐codes and delves into computational efficiency aspects.

Chapter 4 considers the use of NC for managing losses in...

Produkt-Details

• Erscheint lt. Verlag: 28.2.2025
• Sprache: englisch
• Themenwelt: Mathematik / Informatik ► Informatik ► Netzwerke
• Schlagworte: Communication Protocols • Decoding • encoded packets • Finite Fields • information packets • Linear operations • low latency • Multicasting • network efficiency • network infrastructure • network throughput • random linear network coding • small-cell environments
• ISBN-10: 1-394-21729-3 / 1394217293
• ISBN-13: 978-1-394-21729-8 / 9781394217298