Free tools Windows power users keep installed
One-click scans. No signup required.
An IP address often answers two different questions at once: where should packets be sent, and which endpoint is this? Routers need the first answer to forward traffic. Applications often treat the same value as the second. Keeping these roles separate explains why IP addresses change, why MAC addresses are not device identities, why names can outlive addresses, and why some networking designs split the two roles on purpose.
The short answer
Identity says which node, interface, user, or service is involved in a communication. Location says where that thing is attached to the network topology, so that packets can be forwarded toward it. In conventional IP, one address value usually performs both jobs. That is efficient for routing, but it means the address changes whenever the attachment point changes, and it is a weak stand-in for a permanent identity.
![Computer Networking Bible: [3 in 1] The Complete Crash Course to Effectively Design, Implement and Manage Networks. Including Sections on Security, Performance and Scalability](https://m.media-amazon.com/images/I/41H4YJnxKgL._SL160_.jpg)
Four terms that are easy to blur
Identity
An identity is a name or identifier that distinguishes something for a stated purpose. The first step is always to say what is being identified. A node is not the same thing as one of its interfaces, and a user identifier inside an access network identifies a person or account within that domain, not a machine on a particular subnet.
Location (locator)
In this context, location means network topology, not geographic coordinates. A locator is a value that indicates a network attachment or subnet and can be used to route or forward packets toward it. Moving a laptop from an office Wi-Fi network to a home network changes its locator even though the laptop is the same machine.
#1 Best Overall
Address
An address is a value whose meaning depends on the protocol and layer where it is used. An IPv4 or IPv6 address participates in network-layer delivery, and endpoint protocols or applications may also treat it as an identity. A single address should not be reduced to only one role.
Name
A name is a higher-level label, such as a DNS domain name, that can be resolved to one or more current network addresses. A stable name can hide changes in the underlying locator, provided the DNS mapping is updated. Caches can delay that change for some clients until cached records expire.
A worked example: one service, several answers
Consider a service called example.net running on a server. The same request involves four different kinds of value, each answering a different question. The sequence below is the general pattern described in Apple’s developer documentation on addressing and domain names.
- Name. The client wants to reach
example.net. The name is what a person or application keeps and types. - Resolution. DNS returns a currently usable IP address for that name. The name is now mapped to a locator.
- Network-layer forwarding. Routers use the IP address to move packets toward the subnet where the server is attached.
- Local delivery. On the final link, frames are addressed using link-layer information for the next hop, such as the MAC address of the interface that should receive them.
If the server’s connectivity changes and its IP address changes, the operator updates the DNS record. Clients keep using the same name, and the locator behind that name changes. The server’s identity as a service has not changed; only the mapping from name to locator has. Apple’s addressing documentation describes this indirection, although it is an archived document rather than a guide to a current operating system.
The Tool Desk
Outbyte Driver Updater FREEFix the driver behind crashes, sound loss and screen glitchesFind Drivers →Outbyte PC Repair FREERepair Windows errors before they cause bigger problemsFix Now →Where each kind of value sits
The table below uses the layer placement from the ILNP architectural model in RFC 6740, which is an Experimental specification. It is a model for comparison, not a description of how ordinary Internet connections are built.
| Value | Layer in the ILNP model (RFC 6740) | What it answers | Reveals network topology? |
|---|---|---|---|
Fully qualified domain name (e.g. example.net) |
Application | Which service or host name a person or application uses | No, but it resolves to values that may |
| Identifier | Transport | Which node is communicating; defined as a non-topological name | No |
| Locator | Network | Where the node is attached; defined as a topologically bound name for an IP subnetwork | Yes |
| MAC address | Physical interface | Which network interface is on the local link | No, it does not encode a network location |
| IP address in conventional IP | Network (no split) | Both where the node is and which endpoint it is, in one value | Yes |
IP versus MAC, with careful boundaries
An IP address is used at the IP layer and can be IPv4 or IPv6. A MAC address is a link-layer address associated with a network interface and used for delivery on the local link. A host with several interfaces, such as wired and wireless, has several link-layer addresses. Calling a MAC address the permanent identity of the whole device is therefore misleading.
RFC 6155 notes that a MAC address is appropriate for the device using the interface only while the two remain together. Locally assigned MAC addresses are also not guaranteed to be unique outside their administrative domain. A MAC address is useful on a local network, but it is not a global identity for a device.
Why an IP address is not a reliable device identity
An IP address is not automatically durable or globally unique as a device identity. Several common conditions break that assumption.
What’s actually slowing this PC down?
Pick the symptom - the matching free tool is one click away.
- Mobility. A device that moves between networks receives a new locator, so its IP address changes even though the device is the same.
- Dynamic assignment. Addresses assigned by a server or by autoconfiguration can change over time.
- Network address translation (NAT). Multiple internal devices can share one externally visible address. From outside, the address alone does not say which internal device is involved.
- Privacy-oriented addressing. Temporary IPv6 addresses change the interface identifier over time (see the privacy section below).
If you need to tie an external connection to an internal device
- Record the external address, the external port, and the exact time of the connection.
- Check the NAT device’s translation records for that address, port, and time. Whether these records exist, and how long they are kept, depends on the device.
- Map the translated internal address to a device using your own inventory, such as a DHCP lease log or an asset list.
Without the NAT records, the external IP address cannot reliably identify the internal device. This is why a location service may need flow or port context, not only the address.
Rank #4
Why separate identity from location at all?
If a node can keep a stable identity while its network attachment changes, higher layers do not have to treat it as a new node every time it moves. This is the motivation for locator/identifier separation. RFC 6740 describes the intended architectural benefits of ILNP, including mobility and multihoming. Those benefits are design goals of the architecture. They do not mean that an ordinary IP session will survive an address change unless a mechanism designed for that purpose is in use.
Five comparison questions make the distinction usable in practice:
- What object is named? A node, an interface, a user, or a service.
- What is the scope? The local link, an administrative domain, or global routing.
- Does it reveal topology? Whether the value says where something is attached to the network.
- Is it stable? Whether it persists as connectivity changes.
- What is its operational role? Naming, authentication, forwarding, or local delivery.
Two architectures that split the roles differently
ILNP: identifier and locator as separate names
RFC 6740 defines the roles explicitly. Its definitions are: “Identifier: a non-topological name for uniquely identifying a node.” A locator is “a topologically bound name for an IP subnetwork.” This is the clearest standards-level expression of the split. The RFC is labeled Experimental, so it should be presented as an architectural model rather than as a description of how most networks operate today.
Quick wins for a faster PC:
Clear out junk files and repair common Windows errorsFree Scan →Fix the driver behind crashes, sound loss and screen glitchesFind Drivers →Best Value
HIP: a cryptographic identity for hosts
RFC 9063 explains that conventional IP addresses carry a dual locator and endpoint-identifier role. The Host Identity Protocol (HIP) instead uses a cryptographic Host Identity based on a public key. A party that proves possession of the matching private key can demonstrate that identity, regardless of which network address the host currently uses. The identity is therefore not the address, and the address can change while the identity stays the same.
Privacy: stable identifiers make correlation easier
A stable identifier makes activity easier to link together. RFC 8981 explains that changing an IPv6 interface identifier can reduce some correlation over time. It also notes that a stable network prefix can still group activity by network, and a DNS name can remain a recognizable identifier. Temporary addressing therefore limits some exposure, but it does not guarantee anonymity or hide the network attachment.
Quick Recap
What to check when a text uses “identity” or “location”
- Which object is named: the node, the interface, the user, the service, or the address itself?
- Is the value a locator that reveals a network attachment, or a non-topological identifier?
- Is the claim about an IP address on a local link, inside an administrative domain, or across global routing?
- Does the value persist when the device moves or its connectivity changes?
- If the address is shared behind NAT, what extra context is needed to identify the device?
- Is the architecture described in deployed networks, or is it an experimental or architectural model?
Product prices and availability are accurate as of the date/time indicated and are subject to change. Any price and availability information displayed on Amazon at the time of purchase will apply.
