IP Datagram Format: Understanding IP Address Structure

IP Addresses and Datagrams

The IP protocol, which is one of several TCP/IP protocols, provides a hierarchal, hardware-independent addressing system—in other words, it’s the built-in networking software on computers that is needed for delivering data on a complex, routed network like the Internet.

Although it’s the network adapter on a computer that has a unique IP address when it’s connected to the Internet, it’s usually said that the computer has the IP address. That’s because most computers have only one network adapter. The word “host” is commonly used to describe a network device (typically a laptop or desktop computer) that has an IP address.

There are some computers that have multiple network adapters, including a computer that’s acting as a router or a proxy server. Because it has to have more than one network adapter, it also needs more than one IP address.

The Internet isn’t like an old-fashioned switchboard, with a human handling calls and moving information along for you. It’s all digital and networked, and it all happens because of the TCP/IP protocols. Computers, routers and servers figure it all out on the fly, in fractions of a second.

IP address organization.

Once you become familiar with the organization of IP addresses on a network, you can look at an IP address and guess at the location of the host, which is the network or subnet where the host resides.

Part of an IP address acts something like a postal ZIP code: It identifies a general network location. Another part of the IP address is a like house’s street address: In this case, it identifies the computer on the network that the data is going to.

Don’t let that analogy worry you too much as far as your privacy—the IP address does not identify your precise location or identity. It simply tells networks the ultimate destination (which is a computer) for data.

The IP datagram.

Just as the post office and mail carriers deliver letters, data transmitted over the Internet using IP protocols are called messages. They’re also called IP datagrams.

As with all network protocol messages, IP uses a specific format for datagrams. A datagram is divided into and delivered in two segments: the header and the payload.

The header contains addressing and control fields.

The payload carries the actual data to be sent over the network.

In case you’re wondering, a datagram does not have a footer.

An IP header carries quite a bit of information, so it is a fair size in digital terms. It is a minimum of 20 bytes long and can be significantly longer. The TCP/IP software on the source (the sending computer) constructs the IP header. The TCP/IP software at the destination (the receiving computer) uses the information enclosed in the IP header to process the datagram.

So when you check on your IP address using “WhatIsMyIPAddress.com,” you are simply checking your own IP address—if you’re sending an email, it would be the source IP address. If you’re waiting for a reply to an email, you’d be the destination email address.

And in both cases, they’re one in the same.