Section 2: IP Addressing

The Big Picture: Two Kinds of Address

In Section 1 you met the MAC address, the name tag that works on one local network. But the Internet is millions of networks joined together. For a message to cross them, every device needs a second kind of address that says which network it is on and which machine it is on that network.

That is the IP address (Internet Protocol address). Think of a postal address:

Your IP address can change when you move to another network (a café, your office, home). Your MAC address stays with the NIC. That is the main difference.

2.1 IPv4 Structure and Address Classes

What it is: IPv4 (Internet Protocol version 4) is the address system still used by most devices. An IPv4 address is 32 bits long. To make it readable, the 32 bits are cut into 4 bytes and each byte is written as a number from 0 to 255, separated by dots. This is called dotted decimal.

  192      .   168      .   1        .   25
  11000000     10101000     00000001     00011001      ← the same address in bits
  └ 8 bits ┘   └ 8 bits ┘   └ 8 bits ┘   └ 8 bits ┘    = 32 bits in total

There are 2³² = about 4.3 billion possible IPv4 addresses. That sounds like a lot, but the world has more devices than that, which is why NAT (Section 9) and IPv6 (next topic) exist.

Two parts in one address. Every IP address is split into a network part (which street) and a host part (which house). Devices on the same network share the same network part:

   192.168.1.25
   └───┬────┘ └┬┘
   network part  host part     ← where the line goes is decided by the subnet mask (2.3)

Address classes (the old system). Long ago the Internet divided addresses into fixed classes by the first number. You will still meet these names in books and exams:

Today classes are not used. They wasted addresses (a class B network of 65,534 addresses was too big for a company with 300 PCs). The modern system, CIDR (see 2.3), lets the network part be any length. But the "class C" idea, x.x.x.0 with 254 hosts, is still how most home networks look.

Special and private ranges. Some addresses are set aside:

Private versus public. A private address is like an apartment number: it is only meaningful inside the building. A public address is unique on the Internet. Your home devices have private addresses; your router has one public address from your Internet provider. (NAT in Section 9 connects the two.)

Default gateway. To reach any device outside your own network, your PC sends the message to a helper: the default gateway. This is the router's address on your network, for example 192.168.1.1. Think of it as the door to the outside world. Section 8 explains how it works.

See your own addresses (read-only):

ipconfig
ipconfig /all
Get-NetIPConfiguration

2.2 IPv6 Structure and Why It Exists

What it is: IPv6 is the newer version of IP. It was made because IPv4 is running out of addresses. An IPv6 address is 128 bits long (four times longer than IPv4).

How it looks. It is written as eight groups of four hex digits, separated by colons:

   2001:0db8:85a3:0000:0000:8a2e:0370:7334

To save space, two shortcuts are allowed:

1. Drop leading zeros in a group: 0db8 becomes db8.
2. Replace one run of all-zero groups with :: (only once per address).

   2001:db8:85a3::8a2e:370:7334

How many addresses? 2¹²⁸ is about 3.4 × 10³⁸. That is so many that every device on Earth can have a public address, and there is no need for NAT.

Parts of an IPv6 address. The first 64 bits are usually the network (called the prefix), and the last 64 bits identify the device on it.

IPv4 and IPv6 live side by side. Most PCs today have both (this is called dual stack). Windows tries IPv6 first if it works. Both are visible in ipconfig: you will see an "IPv6 Address" and a "Link-local IPv6 Address" starting with fe80::.

Differences that matter:

2.3 Subnetting Basics

What it is: A subnet (sub-network) is a group of devices that share the same network part of the address and can talk to each other directly (without a router). Subnetting is the art of deciding where the network part ends and the host part begins.

The subnet mask. The subnet mask says which bits are the network part. Where the mask has 1 bits, that is the network. Where it has 0 bits, that is the host.

 IP address   192.168.1.25   11000000.10101000.00000001.00011001
 Subnet mask  255.255.255.0  11111111.11111111.11111111.00000000
                             └────── network part ──────┘└ host ┘

Here the mask has 24 ones. So the first 3 bytes (192.168.1) are the network, and the last byte (25) is the host.

CIDR notation (the short way). Instead of writing the whole mask, write the number of 1 bits after a slash. This is called CIDR (Classless Inter-Domain Routing) notation:

The rule for usable hosts: with h host bits there are 2^h addresses, but two are reserved, so you get 2^h − 2 usable ones:

• The all-zeros host is the network address (the name of the network itself).
• The all-ones host is the broadcast address (talk to everyone on the network).

Worked example for 192.168.1.0/24:

Is the other computer on my network? Your PC does this check before every send: apply your mask to your address and to the target address. If the network parts are equal, the target is on the same network, so send directly (using its MAC, Section 3). If they differ, send to the default gateway.

 Me: 192.168.1.25/24     Target: 192.168.1.40   → same network (192.168.1) → send directly
 Me: 192.168.1.25/24     Target: 8.8.8.8        → different network       → send to gateway

Why split a network at all? Smaller networks keep broadcast noise down, let you separate groups (guests, cameras, office PCs) and make rules (firewalls, Section 10) easier.

A tip for the exam. To work out a subnet fast, ask "how many host bits?". Then the block size is 2^host bits. A /26 has 6 host bits, so networks come in blocks of 64: .0, .64, .128, .192.

Summary