What is an IP address?
This definition is for the folks looking at my Subnet Masking definition and summary, but who may not already have a clear understanding of what an IP address is. This isn't meant to be a comprehensive explanation, but just an attempt to fill in some of the concepts that you may be missing in regards to the "structure" of IP addresses. (If you find this too simplistic, fine, I didn't write it for you.)
When you see an IP address such as 184.108.40.206, this is just the easy-for-humans-to-use decimal version of what the computer actually uses: 10111000111011100000111010110100, which is the binary version.
Binary is a numbering system that only uses two digits (0,1) instead of the 10 digits (0-9) you are used to in our decimal numbering system. Since binary is clumsy and error prone for humans, and because of limitations in early computers, programmers began to refer to long sequences of binary digits by breaking them up into groups of 8. They called a single binary digit a "bit," and 8 binary digits a "byte," or a "word." For example, the 32 binary digits above are refered to by breaking them into 4 words, with periods separating the words, like this: 10111000 . 11101110 . 00001110 . 10110100. If that's not confusing enough, we also call these words "octets," since "oct" means eight, and each "word" has 8 binary digits. We also refer to octets as w, x, y, or z when we want to talk about the particular octets, so in the IP address of 220.127.116.11, w is 184, x is 238, y is 14, and z is 180. You can see an example of this in my Subnet Masking definition and in many books.
To do a conversion, you can use the Windows calculator. Find it under Accessories and start it, then click on the View drop down menu, followed by clicking on "Scientific." Type in 184, then click on "Bin," which is short for binary, and the display will convert 184 from decimal to 10111000 in binary. If you want to learn how to convert manually, the best tutorial I've seen is a pamphlet that comes with Transcender's TCP/IP test prep software, but I don't know how you can get it without the software.
So here's the IP address deal: every computer on the whole Internet (including routers) has a unique binary "name" made of 32 binary digits that tell it apart from every other computer on the Internet. If we could use every possible combination of these 32 binary digits, that would be over 4 billion IP addresses. In fact, that's not enough, so the Internet will shift to 64 binary digits for IP address over the next ten years. The 32-bit variation is Internet Protocol version 4 (IPv4) and the 64-bit version is IPv6.
If you have specific questions you need answered, try the Networking Resource Channel at Experts-Exchange.
Copyright © 1999-2006 John Lambert, all rights reserved. John is a premier field engineer with Microsoft. Subnet Masking Articles MENU