IP Addressing and Subnetting Explained

Every device that talks over the Internet Protocol needs an address, and every router needs a way to decide where to send each packet. IP addressing and subnetting solve both problems. Subnetting divides one block of addresses into smaller networks, which keeps broadcast traffic local, separates departments or machines for security, and lets routers forward traffic using short summary routes instead of individual host addresses. It is a core skill for network, telecom and industrial automation engineers alike.

IPv4 address format

IPv4, defined in RFC 791, uses 32 bit addresses. For readability, the 32 bits are split into four 8 bit octets, and each octet is written in decimal from 0 to 255, separated by dots. This is dotted decimal notation.

• 192.168.10.77 in binary is 11000000.10101000.00001010.01001101.
• The total address space is 2³² addresses, about 4.3 billion.

Each address has two parts:

• Network part (prefix): identifies the network or subnet. All devices on the same subnet share it.
• Host part: identifies one interface on that subnet.

IP sits at the network layer, layer 3, of the models compared in OSI vs TCP/IP. Switches forward frames inside a subnet using MAC addresses, while routers forward packets between subnets using IP addresses, as explained in router vs switch.

Subnet masks and CIDR prefixes

A subnet mask is a 32 bit value with 1s in the network bit positions and 0s in the host positions. A device finds its network address by a bitwise AND of its address and the mask.

Classless Inter-Domain Routing (CIDR), described in RFC 4632, writes the mask as a prefix length: the number of leading 1 bits. So 255.255.255.0 is /24 and 255.255.255.192 is /26. CIDR replaced the original classful scheme, in which the first bits of an address fixed the network size (class A /8, class B /16, class C /24). Classful sizes wasted addresses, and CIDR allows any prefix length.

For a prefix length n, the number of host bits is 32 minus n, and:

$$\text{Addresses per subnet} = 2^{32-n} \qquad \text{Usable hosts} = 2^{32-n} - 2$$

The two addresses removed are the network address (all host bits 0) and the broadcast address (all host bits 1). RFC 3021 makes an exception for /31 subnets on point to point links, where both addresses are used.

Download diagram (WebP)

Private, public and special IPv4 addresses

RFC 1918 reserves three ranges for private networks. They can be reused inside any organisation and are not routed on the public internet:

• 10.0.0.0/8 (10.0.0.0 to 10.255.255.255)
• 172.16.0.0/12 (172.16.0.0 to 172.31.255.255)
• 192.168.0.0/16 (192.168.0.0 to 192.168.255.255)

Devices with private addresses reach the internet through network address translation (NAT) on a router or firewall, which maps them to a public address. Other special blocks include 127.0.0.0/8 for loopback (127.0.0.1 is the local host) and 169.254.0.0/16 for link local addresses that a host assigns itself when no DHCP server answers.

How IP addressing and subnetting work step by step

To find the subnet details for any address and prefix:

1. Find the host bits: 32 minus the prefix length.
2. Find the block size: 2 to the power of the host bits, counted in the octet where the prefix ends. For /26 the block size is 64 in the fourth octet.
3. Find the network address: the largest multiple of the block size that is not greater than the address in that octet.
4. Find the broadcast address: network address plus block size minus 1.
5. Find the usable range: from network plus 1 to broadcast minus 1.

When dividing a block into equal subnets, borrow enough host bits to make the number of subnets needed. Borrowing b bits creates 2 to the power b subnets. Variable length subnet masking (VLSM) goes further and gives each subnet only the size it needs, for example /30 or /31 for router links and /24 for a user LAN.

IPv6 addressing basics

IPv4 addresses ran out at the regional registries, which is why IPv6 was developed. IPv6 addresses, defined in RFC 4291, are 128 bits long and written as eight groups of four hexadecimal digits separated by colons, for example 2001:db8:0:1::10. Leading zeros in a group can be dropped, and one run of all zero groups can be replaced by a double colon.

In IPv6 a normal LAN subnet is a /64: the first 64 bits are the network prefix and the last 64 bits are the interface identifier. Sites typically receive a larger prefix, such as a /48 or /56, and create /64 subnets from the bits in between. IPv6 has no broadcast address; it uses multicast instead.

Common subnetting mistakes

• Assigning the network or broadcast address to a device.
• Using a mask on a host that does not match the router interface, so the host cannot reach its gateway.
• Overlapping subnets in different sites that must later be connected by VPN.
• Forgetting to plan address space for future growth, management networks and point to point links.

Correct addressing is the foundation for everything above it, from how HTTP works on a web server to industrial Ethernet networks in a plant.