This article presents an overview of several key improvements offered by the Internet Protocol version 6 (IPv6) over current Internet Protocol version 4 (IPv4). The topics covered include IPv6 addressing and routing concepts, changes to the minimum IPv6 packet size, flows, and traffic classes, the neighbor discovery and node autoconfiguration mechanisms, and an overview of mobile IPv6 and the network security architecture. Transition mechanisms, such as dual stacks and the 6bone, are also discussed. The 6bone is a virtual network that is used to help test and facilitate the development of IPv6. Key conce ts associated with the 6bone, such as setup requirements, IPv6 DNS support, anltunnel mechanics, are also presented he rapid growth of the current Internet, which operates using the Internet Protocol version 4 (IPv4), has created a number of problems for the administration and operation of the global network. These problems include the decreasing number of available IPv4 addresses for network nodes, and the rapid growth of memory and performance requirements for network routers. While changes to IPv4 have extended the life of the current Internet, these changes tend to create new problems and require a significant amount of overhead for network administration. The Internet Protocol version 6 (IPv6)l has been designed to support these extensions, and more, without creating the additional problems. An excellent discussion of the requirements and background €or Ipil6 can be found in [l].This article provides a detailed introduction to the addressing specifications, the prefix-based routing scheme, changes to the IPv6 minimum packet size, flows, and traffic classes, neighbor discovery and address autoconfiguration, mobile IPv6, and the IP security architecture. Additionally, ongoing IPv6 testing and deployment efforts are reviewed. Neither the IPv6 packet format nor the Internet Control Message Protocol version 6 (ICMPv6) is discussed in this article; however, a detailed discussion of these and other topics may be found in For the curious, IPv5 is the experimental Stream Transport Ilprotocol.[l-61. This article differs from other articles by providing more up-to-date information, and delves into different technical features of IPv6.
AddressingNetwork addresses serve two purposes: first, to uniquely identify an interface, and second, to aid routing by identifying "where" an interface is on the network. The 32-bit IPv4 address is divided into different classes, consisting of fixed boundaries between the identification of each network and the identification of each node. For example, a class B address has a theoretical possible range of 65,536 addresses, and a class C address has a possible range of 256 addresses. Fixedsized addresses cause problems because most sites will need more than 256 addresses and fewer than 65,536 addresses. An IPv6 address is 128 bits in length, and the boundary between the transit network and site network subdivisions is not fixed, so networks can be sized accordingly. This allows a more flex...
