UK IPv6 Taskforce > Why a new Internet Protocol?

Why a new Internet Protocol?

The world population is now about 6.2 billion people with 1.5 billion projected users on the Internet during 2004.

The success of the Internet leaves us today with probably less than a third of the Internet (IPv4) addresses still available out of the 4.3 billion theoretical maximum.

In the face of conservative projected demand of three to four billion Internet addresses over the next five years there is an address shortage on the horizon. Internet cannot be for everyone if we continue to depend on IPv4.

Methods of Internet address space conservation are widely used but they may yet fail to manage this growth. All such methods are deployed at some expense to application flexibility and user security. They also add significant expense and complexity for users and networks. Inevitably we will reach address space exhaustion. It is just a question of time.

The threat of address exhaustion is likely to be even more urgent than the above figures suggest as the methods of IP address allocation and aggregation implies that many fewer than the theoretical 4.3 billion possible addresses are in practice available.

IPv6, the next generation Internet Protocol, has been developed from the ground up to scale the address space to the huge success that the Internet has become. IPv6 takes the best of the intended open, end-to-end features of the IPv4 protocol whilst ensuring sufficient addresses for the foreseeable future to restore end-to-end capability where it has already been broken due to address shortages. (IPv6 has approximately 340 billion, billion, billion, billion (34x10³⁸) addresses available.)

One of the key features of IPv6 is that it can be used alongside IPv4. This allows a gradual migration over some years from IPv4 to IPv6. The deployment of IPv6 will not require a Big Bang migration date as was experienced in 1983 when the Internet then sized at only 400 hosts moved from NCP to IP in a single step. Internet today is considered far too large to make a single transition event to IPv6 desirable.

However gradual transition methods to IPv6 require a large number of IPv4 addresses. The transition to IPv6 will further exacerbate the address space shortage until IPv6 is widely deployed natively.

As the scale of future demand for Internet services is difficult to predict and the cost of IPv4 space due to shortages is likely to escalate it is strategically important for all Internet users, content and application providers and networks to start to support and migrate to IPv6 during the 2003 to 2006 timeframe. Leaving it later could require a more drastic and expensive migration policy.