William is a frequent contributor to this magazine. His most recent book is Local and Metropolitan Area Networks, Fifth Edition (Prentice Hall, 1996). He can be contacted at ws@shore.net.
The Internet Protocol (IP) has been the foundation of the Internet and virtually all multivendor private internetworks. This protocol is reaching the end of its useflife and a new protocol, known as IP Version 6 (IPv6), has been defined to ultimately replace IP.
The driving motivation for the adoption of a new version of IP was the limitation imposed by the 32-bit address field in IPv4. IP also is a very old protocol and new requirements in the areas of security, routing flexibility, and traffic support have developed. In response to these needs, the Internet Engineering Task Force (IETF) issued a call for proposals for a next-generation IP (IPng) in July 1992. A number of proposals were received, and by 1994, the final design for IPng emerged. A major milestone was reached with the publication of the Recommendation for the IP Next Generation Protocol (RFC 1752) in January 1995. RFC 1752 outlines the requirements for IPng, specifies the PDU formats, and highlights the IPng approach in addressing, routing, and security. A number of other Internet documents defined details of the protocol, now officially called IPv6. These include an overall specification of IPv6 (RFC 1883), an RFC discussing the Flow Label in the IPv6 header (RFC 1809), several RFCs dealing with addressing aspects of IPv6 (RFC 1884, RFC 1886, RFC 1887), and five security-related RFCs (RFC 1825-RFC 1829).
IPv6 provides a number of enhancements over IPv4: the use of 128-bit addresses rather than 32-bit addresses, a new type of address called "anycast," an address autoconfiguration scheme, an improved mechanism for specifying options, a resource allocation scheme based on flow labeling, and security features.
IPng: Internet Protocol Next Generation, edited by Scott Bradner and Allison Mankin, and IPv6: The New Internet Protocol, by Christina Huitema, should satisfy any interests you have in the new Internet protocol and direction.
IPng: Internet Protocol Next Generation is a collection of chapters edited by the two cochairs of the IETF/IPng process. This book was finished somewhat ahead of the standards process and, thus, is not heavy on technical details of actual specifications. Instead, the aim of the book is to explain the rationale behind the format and numerous features in IPv6 and to discuss the various applications and networking technologies that IPv6 can potentially support. The book is relatively nontechnical and is accessible to anyone with a basic understanding of networking and communications.
Following a brief introduction, part 2 of the IPng book describes the process by which IPv6 was developed within the IETF. Part 3 examines the expected lifetime of the 32-bit address space, which was the issue that triggered the IPv6 process. The next two parts, consisting of about 140 pages, are the most important sections of the book. In December 1993, the IPng workgroup issued a call for white papers to help define the requirements for, and the features of, the next-generation IP. In response, 21 papers were received. These papers, in modified form, are reprinted in the book. Part 4 contains papers addressing requirements such as including IPng in large corporate networks; supporting ATM with IPng; and separate chapters on requirements for the Navy, the electric power industry, the cellular industry, and the cable TV industry. Part 5 contains papers on suggested features; topics include addressing, routing, and security. Together, these two parts constitute a unique survey of internetworking technology and applications, and the requirements that these impose on the Internet protocol.
The next part of the book briefly describes the various proposals that were considered for IPng. The remainder, about 50 pages, is an overview of IPv6 itself.
Christina Huitema, the author of IPv6: The New Internet Protocol, is the former chair of the Internet Activities Board and is active in the IPng effort. This book is a straightforward, technical description of the various RFCs that make up the IPv6 specification. Interestingly, the competition for this book is not the Bradner/Mankin book, but the RFCs themselves, which are clearly written and well organized. However, Huitema provides additional explanations of various features and operations of the protocol.
The organization of Huitema's book closely follows the breakdown of the IPv6 specification into multiple RFCs. After an introduction, chapter 2 provides an overview of the format of IPv6, which consists of a base IPv6 header and a number of extension headers. The objective was to keep the base header relatively small and of fixed size. (By relegating many features to optional extension headers, the processing burden on routers is minimized.) Next, four of the optional headers are examined: the routing header, the fragmentation header, the destination header, and the hop-by-hop options headers. Huitema explains the role of each and the relationship between them.
Chapter 3 deals with address-related issues. Huitema first discusses the possible ways in which the 128-bit address can be formatted and the assignments that have been made to date. The very important issue of how address format and assignment relates to routing is detailed. Chapter 4 is devoted to autoconfiguration: techniques for automatically discovering and assigning IPv6 addresses. This is one of the best and most useful chapters in the book, because it provides much detail not found in the RFCs. Chapter 5 treats the security-related features of IPv6: Huitema provides a clear explanation of how confidentiality and authentication are implemented in IPv6 (the same options also may be used in the current IP). Chapter 6 discusses the specific issue of defining and managing flows using IPv6. Finally, chapter 7 discusses transition strategies. The book is easy to follow, up to date, and well written. Each chapter ends with a section entitled, "Points of Controversy," which provides some background for the design decision and assesses alternative approaches that might have been used.
In a nutshell, the Bradner/Mankin book is mainly about the "why" of IPv6, while the Huitema book is mainly about the "how" of IPv6, although each does address both themes. If you are interested in IPv6, you should read both books.