The Internet Engineering Task Force (IETF) sets many of the policies related to the Internet and communication via the Internet Protocol.
IETF standards are published as “Request For Comments” or “RFCs.”
RFC 1918 sought to address the eventual exhaustion of IP addresses, which we have since experienced. It introduced and standardized the concept of “private” and “globally unique” IP addresses.
RFC 1918 defined the following ranges as “private IP addresses:”
10.0.0.0 – 10.255.255.255 (10/8 prefix)
172.16.0.0 – 172.31.255.255 (172.16/12 prefix)
192.168.0.0 – 192.168.255.255 (192.168/16 prefix)
That is, it determined that any address within these ranges would not be routable on the public Internet.
Globally unique IP addresses are the only addresses routable via the public Internet.
It is very interesting to note, if you read the whole text of RFC 1918, that NAT is never mentioned, as it had not been commercially developed at the time of the RFC’s release.
The conservation of globally unique IP addresses comes from the assumption that organizations that need many addresses in a large network could use the private addresses throughout their internal network without concern for efficiency and subsequently use NAT at the network edge. NAT enables many internal IP addresses to use only one or just a few globally unique IP address to communicate with parties via the Internet.
IP Version 6 allows for many more globally unique IP addresses than IPV4.
RFC 1918 is the reason merging corporations often have network addressing schemes that conflict with each other. In other words, the same addresses are used within each private network, as RFC 1918 intended. One network usually has to undergo a complete re-addressing in order for full access between infrastructures to occur.