Network address translator traversal is a computer networking technique of establishing and maintaining connections across gateways that implement network address translation (NAT). NAT breaks the principle of end-to-end connectivity originally envisioned in the design of the Internet.
NAT traversal techniques are required for many network applications, such as peer-to-peer file sharing and Voice over IP.
NAT devices are commonly used to alleviate IPv4 address exhaustion by allowing the use of private IP addresses on private networks behind routers with a single public IP address facing the public Internet. The internal network devices communicate with hosts on the external network by changing the source address of outgoing requests to that of the NAT device and relaying replies back to the originating device.
This leaves the internal network ill-suited for hosting servers, as the NAT device has no automatic method of determining the internal host for which incoming packets are destined. This is not a problem for general web access and email. However, applications such as peer-to-peer file sharing, VoIP services, and video game consoles require clients to be servers as well. Incoming requests cannot be easily correlated to the proper internal host. Furthermore, many of these types of services carry IP address and port number information in the application data, potentially requiring substitution with deep packet inspection.
Network address translation technologies are not standardized. As a result, the methods used for NAT traversal are often proprietary and poorly documented. Many traversal techniques require assistance from servers outside of the masqueraded network. Some methods use the server only when establishing the connection, while others are based on relaying all data through it, which increases the bandwidth requirements and latency, detrimental to real-time voice and video communications.