Simple definition
PAT (Port Address Translation) is a form of NAT that allows many internal connections to share one public IP address by distinguishing sessions with port numbers.
Technical definition
When an internal host opens a session to the Internet, the NAT device replaces the private source address with a public address and can also select a translated source port. It keeps a state table that maps the internal tuple to the translated tuple so return traffic can be delivered to the correct host.
How it works / role
Its role is to multiplex many private sessions behind one or more public addresses. On some platforms this mechanism is commonly described as NAT overload.
What is it used for?
Provide Internet access to many private hosts through a single public IPv4 address while preserving a distinct mapping for every session.
Practical example
Hosts 192.168.1.10 and 192.168.1.20 connect to the same website. The router uses the same public IP address but different translated source ports, allowing each response to return to the correct host.
Common issues
- Port exhaustion under a very large number of sessions
- NAT timeout is too short for a specific application
- Asymmetric routing prevents return traffic from crossing the same stateful device
- Confusion between outbound PAT and inbound port forwarding
Key takeaway: PAT conserves public IPv4 addresses but depends on state tracking, so session and port capacity are important sizing factors.