Stateful NAT64 Overview
Network Address Translation (NAT) is a mechanism for concealing a set of host addresses on a private network behind a pool of public addresses. NAT64 is a related technology that allows IPv6-only clients to contact IPv4 servers using Unicast UDP, TCP, or ICMP.
A public IPv4 address is shared among several IPv6-only clients. To accomplish this, NAT64 translates incoming IPv6 packets into IPv4 and vice versa. This solution allows ISPs to move to an IPv6 network while simultaneously handling IPv4 address depletion.
When stateful NAT64 is used in conjunction with DNS64, changes are not usually required in the IPv6 client or the IPv4 server. (DNS64 is not covered in this document. For details, see draft-ietf-behave-dns64-12, DNS64: DNS extensions for Network Address Translation from IPv6 Clients to IPv4 Servers and RFC6147 DNS extensions for Network Address Translation from IPv6 Clients to IPv4 Servers.)
Problems Resolved by NAT64
NAT64 solves the problem of IPv6 clients initiating connections to IPv4 servers. A cornerstone of the architecture is that NAT64 is easily deployable because it does not require changes to either the IPv6 client or the IPv4 server.
NAT64 today works much like a symmetric Network Address Port Translation (NAPT44). In NAPT, both the original source address and the source port are translated. The translated address and port are picked up from the corresponding NAT pool.
IPv6 packets originated by IPv6 clients are transparently translated to IPv4 by a NAT64 device. The device performs the packet header translation according to the IP/ICMP translation algorithm, translating the IPv4 addresses of IPv4 hosts to and from IPv6 addresses and assigning a prefix to the stateful NAT64 function for the translation. For basic functionality, you deploy the stateful NAT64 function in the NAT64 device, in addition to a few DNS64-enabled name servers accessible to the IPv6-only hosts. (For an analysis of the application scenarios, see draft-ietf-behave-v6v4-framework-10 and RFC 6144 Framework for IPv4/IPv6 Translation.)
Configuring Address Translation
To perform IPv6-to-IPv4 translation, the NAT64 function binds an IPv6 address and port (called an IPv6 transport address) to an IPv4 address and port (called an IPv4 transport address). The translation maps the IPv6 transport addresses to IPv4 transport addresses and vice-versa. To create these mappings, the router running NAT64 has two pools of addresses: an IPv6 address pool (to represent IPv4 addresses in the IPv6 network) and an IPv4 address pool (to represent IPv6 addresses in the IPv4 network).
The IPv6 address pool is an IPv6 prefix assigned to the translator
itself. This release supports only IPv6 /96 prefixes, both
the well-known prefix (64:FF9B::/96) and arbitrary /96 prefixes. Due to the abundance of IPv6 address space, the
prefix can be equal to or even bigger than the whole IPv4 address
space. This allows each IPv4 address to be mapped into a different
IPv6 address by simply concatenating the prefix with the IPv4 address
being mapped and a suffix (that is, an IPv4 address X is mapped into
the IPv6 address Prefix:X:Suffix).
The IPv4 address pool is a set of IPv4 addresses, normally a small prefix assigned by the local administrator. Because IPv4 address space is a scarce resource, the IPv4 address pool is small and typically not sufficient to establish permanent one-to-one mappings with IPv6 addresses. Thus, except for manually created mappings, mappings using the IPv4 address pool are created and released dynamically. Moreover, because of the IPv4 address scarcity, the mapping takes place between transport addresses rather than actual addresses.
For additional details about the translation algorithms, see the following documents:
draft-ietf-behave-v6v4-xlate-23, IP/ICMP Translation Algorithm.
RFC 6145, IP/ICMP Translation Algorithm.
draft-ietf-behave-address-format-10, IPv6 Addressing of IPv4/IPv6 Translators.
RFC 6052, IPv6 Addressing of IPv4/IPv6 Translators.