2. Motivation

With the proliferation of TCP/IP technology worldwide, including outside the Internet itself, an increasing number of non-connected enterprises use this technology and its addressing capabilities for sole intra-enterprise communications, without any intention to ever directly connect to other enterprises or the Internet itself.

The Internet has grown beyond anyone's expectations. Sustained exponential growth continues to introduce new challenges. One challenge is a concern within the community that globally unique address space will be exhausted. A separate and far more pressing concern is that the amount of routing overhead will grow beyond the capabilities of Internet Service Providers. Efforts are in progress to address these concerns through various approaches. These efforts include development of a new generation of IP (IPng), and improvements to current usage of IP.

The IANA's currently recommended address allocation strategy for the existing IP is to have Internet Service Providers (ISPs) allocate addresses to the enterprises they serve. This provides for address allocation that incorporates routing layer aggregation, to keep routing overhead from growing at rates that will not be sustainable by the Internet's service providers as the number of Internet hosts grows.

One positive side-effect of this strategy is that the rate of allocation of IP network numbers has been slowed significantly relative to the number of hosts required on the Internet. However, many enterprises use TCP/IP heavily outside their connection to the Internet, and the intent to connect to the Internet at best is ambiguous at many such sites.

Such enterprises need a sizeable chunk of address space, since they may have a considerable number of hosts, and have no intention or plan to connect to the Internet, or may only connect in the distant future. In fact, even when a decision to connect is made, the timing of such a connection may be impossible to predict that far in advance, and it may even be impossible to determine if reserving IP addresses will be needed.

For networks that have a relatively small number of hosts, the administrative overhead of obtaining unique IP addresses may not seem to be warranted. On the other hand, even large enterprises may be unwilling to expose their internal topology, even if the cost of doing so may be low, for privacy or security considerations.

It is advantageous to have a clearly distinguishable range of addresses for such hosts. This would allow address allocators to clearly understand the scope and usage of private addresses, enabling them to make more precise estimates of future address requirements. It would also make router configuration easier to manage traffic from inside an enterprise to the external Internet.