← All posts
6 min read

IP Addresses: Locations for Routing, Not Permanent Identities

#technology#computer-science#ip-addresses#networking
📑 On this page

Routers need a scalable way to decide where packets should go. They cannot identify destinations only by a person's name or by the serial number of a physical computer.

Internet Protocol uses addresses associated with network interfaces.

An IP address identifies a logical network location for routing at a particular time. It is not a permanent identity for a person or device.

The address often changes when a device joins another network, reconnects, or uses privacy features.

Address plus prefix

An IP address is interpreted with a network prefix.

In IPv4, a notation such as:

192.168.1.42/24

means the first 24 bits identify the network prefix. The remaining bits distinguish addresses within that network.

Devices compare prefixes to decide whether a destination appears local.

  • Local destination: send directly across the local link.
  • Non-local destination: send to a router, commonly the default gateway.

The prefix is as important as the address. The same address with a different prefix length produces a different network boundary.

Addresses belong to interfaces

A laptop can have several IP addresses:

  • Wi-Fi address
  • Ethernet address
  • VPN address
  • Loopback address
  • IPv4 and IPv6 addresses

The operating system chooses an outgoing interface and source address based on routes and policy.

Saying "the computer's IP" is convenient but simplified. Addresses attach to interfaces and network contexts.

Dynamic assignment

Most home devices receive IPv4 configuration through DHCP.

A DHCP service can provide:

  • IP address
  • Prefix or subnet mask
  • Default gateway
  • DNS server
  • Lease duration

The address is leased, not permanently owned by the device.

When the lease expires or the device joins another network, it can receive a different address.

IPv6 also supports automatic configuration through router advertisements and DHCPv6.

Static addresses

A static address is configured to remain consistent.

Servers, routers, printers, and infrastructure may need stable addressing.

Static configuration must avoid conflicts and use the correct prefix, gateway, and DNS information.

Home routers can also reserve a predictable DHCP lease for a device. This centralizes address management while keeping the device configured automatically.

A static address does not mean public or globally reachable. It can be static and private.

Private IPv4 addresses

Specific IPv4 ranges are reserved for private networks, including:

10.0.0.0/8
172.16.0.0/12
192.168.0.0/16

These addresses are not routed across the public internet.

Many homes can use 192.168.1.42 simultaneously because each address exists within a separate private network.

The home router commonly translates private connections to a public address using network address translation.

Private addressing conserves global IPv4 space but adds translation and inbound-connectivity complexity.

Public addresses

A public IP address is globally routable under internet routing policy.

An internet provider may assign one to a home router. Cloud servers can receive public addresses or sit behind load balancers and translation gateways.

Public does not mean that every port is reachable. Firewalls, routing, provider policy, and application listeners still determine access.

Conversely, a service without a direct public address can be exposed through a reverse proxy, tunnel, or port-forwarding rule.

A concrete example: visiting a website

Suppose a laptop has:

address: 192.168.1.42/24
gateway: 192.168.1.1

The website destination is outside 192.168.1.0/24.

  1. The laptop sends the packet toward gateway 192.168.1.1.
  2. The router translates or forwards it through the provider.
  3. Internet routers use the destination prefix.
  4. Replies return toward the router's public address.
  5. Translation state directs them back to 192.168.1.42.

The website usually sees a public address associated with the household or provider path, not the laptop's private address.

Loopback addresses

Loopback represents the local host itself.

Common examples:

127.0.0.1
::1

The hostname localhost often resolves to loopback.

Traffic sent to loopback does not leave the machine. Developers use it to test local services.

A service listening only on loopback is not directly reachable from other devices, which can be a useful security boundary.

Link-local addresses work only on the directly connected local link.

IPv6 automatically uses link-local addressing for essential neighbor and router communication.

IPv4 can self-assign addresses in 169.254.0.0/16 when normal configuration fails. Seeing such an address can indicate that the device could not reach DHCP.

Routers do not forward ordinary link-local traffic across networks.

Addresses and identity

An IP address is weak identity evidence.

Several users can share one public address through:

  • Home NAT
  • Company gateway
  • Carrier-grade NAT
  • Proxy

One user can appear from several addresses while moving among Wi-Fi, cellular, VPN, and cloud services.

Addresses can also be spoofed in protocols that lack a return-path handshake.

Authentication should use credentials, keys, certificates, or other stronger mechanisms.

Address conflicts

If two devices use the same address on one local network, communication becomes unreliable.

Symptoms include:

  • Intermittent access
  • Wrong device receiving traffic
  • Duplicate-address warnings
  • Connections failing after another device starts

Automatic configuration and conflict-detection mechanisms reduce this risk.

When assigning static addresses manually, keep them outside unmanaged dynamic pools or create reservations.

Common misunderstandings

"An IP address identifies one person"

Addresses can be shared, reassigned, translated, proxied, or changed.

"A public IP means the computer is fully exposed"

Reachability also requires routing, firewall permission, and a listening service.

"A private IP is secret"

Private ranges are publicly documented. Their value comes from non-global routing, not secrecy.

"Every device has only one IP address"

A device can have multiple interfaces, address families, and temporary addresses.

Knowledge check

1. What does /24 mean in an IPv4 address?

The first 24 bits are the network prefix used for routing and local-network decisions.

2. Why can many homes use 192.168.1.42?

It belongs to a private range and exists independently within each private network.

3. What does a 169.254.x.x IPv4 address often indicate?

The device self-assigned a link-local address, often because normal DHCP configuration was unavailable.

4. Why is an IP address poor authentication?

It can be shared, changed, translated, proxied, reassigned, or sometimes spoofed.

The one idea to remember

An IP address describes a routable network location for an interface, not a permanent human or hardware identity.

Prefixes organize addresses into networks, configuration can be dynamic or static, and scope determines whether an address is local, private, or globally routable.

Next, we will compare the two IP address families operating across today's internet.