iTechGuides is reader-supported. When you buy through links on our site, we may earn an affiliate commission. As an Amazon Associate I earn from qualifying purchases. Learn more
An IP address usually answers two questions at once: which endpoint is involved, and where that endpoint is attached to the network. Because one value does both jobs, a changed address can look like a changed device, and a device’s address can look like its permanent identity. Separating the two roles explains how mobility and multihoming work, why privacy is harder than it looks, and why names, IP addresses and MAC addresses are not interchangeable.
Identity and location answer different questions
Identity asks which thing is being referred to. Location asks where traffic has to go so that it reaches that thing. Both questions matter in every network, but ordinary IP addressing answers them with the same number, so the distinction is easy to miss.
![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)
Identity
An identity is a name or identifier that distinguishes a node, endpoint, interface, user or device 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 network access identifier can identify a user inside a particular domain without identifying a machine at all.
Recommended Free Tools
Location, or locator
In networking, “location” means a position in the 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. An IP address that lets routers pick the next hop is doing locator work.
#1 Best Overall
Address
An address is a value whose meaning depends on the protocol and the layer. In conventional IP, an address takes part in network-layer delivery, and endpoint protocols or applications may also treat the same value as the identity of a host. It should not be reduced to only one role.
Name
A name is a higher-level label, such as a DNS domain name, that resolves to one or more current network addresses. A stable name can hide changes in the underlying address, provided the DNS mapping is updated.
Following one connection through the layers
The clearest way to see the separation is to follow a request for a server called example.net from start to finish. The domain name is what a person or application keeps. The network addresses are what the network uses at each hop.
- The client asks DNS for the name
example.net. The answer is a currently usable IP address. The name is the identity the user recognises; the address is only a current locator. - The client sends IP packets addressed to that address. Routers use network-layer addressing to move each packet toward the network that holds the server.
- On each local link, frames are delivered to the next hop using link-layer addresses. The MAC address names an interface on that link, not the server as a whole.
- If the server moves to another network and its address changes, the DNS record is updated. Clients that re-resolve the name reach the server at its new locator. Clients holding a cached answer keep using the old address until the cache expires, which depends on the record’s time-to-live value.
Source for the addressing concepts: Apple Developer Documentation, Addressing Schemes and Domain Names. That page is an archived explanatory reference, not a guide to a current operating system.
IP and MAC addresses: where the boundaries lie
An IP address is used at the IP layer and can be IPv4 or IPv6. A MAC address is a link-layer address tied to a network interface and used for local delivery. A host with several interfaces, such as a wired and a wireless adapter, has a separate link-layer address for each one.
| Value | Layer or role | Scope | Says where it is attached? | Stable identity for a device? |
|---|---|---|---|---|
| DNS name (FQDN) | Application-facing label | Global namespace | No, it resolves to addresses | Stable as a label, but it names a service or host, not a physical machine |
| IP address (IPv4 or IPv6) | Network layer | Routing domain or global routing | Yes, it reflects the attachment and subnet | Not automatically; it can change and can be shared behind NAT |
| MAC address | Link layer, per interface | Local link or administrative domain | Only on the local link | Tied to an interface; not a universal identity for the whole device |
Why one interface address is not a device identity
RFC 6155 states that a MAC address is appropriate for the device using the interface as long as the two stay together. It also notes that locally assigned MAC addresses are not guaranteed unique outside their administrative domain. Treating a MAC address as the permanent name of a computer therefore overstates what the value guarantees. The same RFC discusses location services that must use more than an IP address. Behind NAT, several devices can share one externally visible address, so a location service may need port or flow context and knowledge of the NAT mapping to tell an internal device apart. An IP address alone is not a durable, globally unique device identity. (RFC 6155)
Rank #4
Why separate identity from location?
If the identity of an endpoint is kept stable while its locator is updated, a node can change where it attaches to the network without every higher layer treating it as a new node. That is the core motivation for mobility and multihoming designs. The separation is a design choice, not a description of ordinary IP connections, which combine both roles in one address.
Outdated Drivers Are Slowing You Down
One free scan finds every outdated or missing driver and matches the right update for your exact hardware.Free scan · exact hardware matchWindows Errors? Fix Them Before They Spread
Repair common Windows errors and clear accumulated junk for a smoother, more stable PC - no reinstall needed.Free scan · no reinstallIdentifier/locator separation in ILNP
RFC 6740 describes the Identifier-Locator Network Protocol (ILNP), an architecture that gives each role its own value. It defines an identifier as “a non-topological name for uniquely identifying a node” and a locator as “a topologically bound name for an IP subnetwork” (RFC 6740, section 2.1). Its layered view places the values as follows:
Best Value
| Layer in the ILNP model | Value named in RFC 6740 | Question it answers |
|---|---|---|
| Application | Fully qualified domain name (FQDN) | Which service or host name is wanted? |
| Transport | Identifier | Which node is this session for? |
| Network | Locator | Where in the topology is that node attached? |
| Physical interface | MAC address | Which interface is delivering the frame locally? |
RFC 6740 is labelled Experimental. Present ILNP as an architectural model that makes the roles explicit, not as a claim that everyday Internet connections already use a pure split. (RFC 6740)
Host Identity Protocol: a cryptographic identity
RFC 9063 explains that conventional IP addresses carry a dual locator and endpoint-identifier role. The Host Identity Protocol (HIP) takes a different route. Its Host Identity is a cryptographic identity based on a public key, and the holder of the matching private key proves that identity. The identity is therefore not derived from where the host is attached. (RFC 9063)
What the separation does and does not promise
The architectural benefit is that the identity can stay constant while the locator changes. It does not mean that an ordinary IP session survives an address change. Surviving that change needs a mechanism designed for it, such as the ones these architectures describe.
The Tool Desk
Outbyte PC Repair FREERepair Windows errors before they cause bigger problemsFix Now →Outbyte Driver Updater FREEScan for outdated or missing drivers - takes under a minuteDriver Scan →Privacy: stable values make correlation easy
Stable identifiers make activity easier to link together. RFC 8981 explains that changing an IPv6 interface identifier reduces some of that correlation. A stable network prefix can still group activity by network, and a DNS name can remain a recognisable identifier. Temporary addressing therefore limits some exposure. It does not make a user anonymous, and it does not hide the network the device is attached to. (RFC 8981)
Quick Recap
Key points to keep
- Identity and location are different functions, even when one IP address value serves both.
- Conventional IP addresses combine the locator and endpoint-identifier roles, and how they behave depends on the protocol layer and the application.
- DNS names can stay stable while the address records behind them change.
- MAC addresses belong to interfaces on a local link. They are not universal identities for an entire device.
- NAT, temporary addressing and multiple interfaces make claims of a single permanent, unique address unreliable.
- Identifier/locator separation is an architectural approach, and the RFCs above describe it as such. It is not a description of every deployed IP network.
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.
