CertiGuide to Network+  9  Chapter 0101:  TCP/IP

II  TCP/IP Addressing Overview 1 Class A

As we noted when discussing protocols, all devices communicating via TCP/IP have a unique address in a format specific to TCP/IP. With the current IP standard (sometimes called Ipv4), the address must be a 32-bit address. At the native level, it is 32 0’s and 1’s. To make it simpler for humans to read, it is broken down to 4 different numbers that range from 0 to 255, with each number representing the value of one octet (8 bits) of the 32-bit address. (Recall that 255 is the largest binary number that can be held in 8 bits.) The sequence of numbers is separated by the period (dot).

To the left side the numbers are the unique number for the network or LAN. These numbers are referred to as the “network number” or “subnet number”. The numbers on the right are reserved for unique numbers for the host (computer or other device). The class of the network address determines the break point between the bits used for the number and host number. Because you know you are restricted to a total of 32 bits to hold all of this information, there’s a tradeoff involved in deciding how much of the IP address is used to hold the network number, and how much is used to hold the host number.

In general, the longer the network number, the shorter the host number portion. This means that if you have a long network number, there are fewer possible host addresses that you can define, and thus fewer devices that you can have on your network. Given this, you might think, “Well, then I want a really short network address, because I want LOTS and LOTS of hosts on my network.” That’s fine in theory, but network addresses for networks that will be connected to the Internet are a bit like beachfront property in Florida – since every network connected to the Internet has to have a unique network ID, and the Internet’s been around for a while, odds are “the best (shortest) network addresses are already taken.”

Your ISP has to approve your network number (in most cases today, they’ll just TELL you which one you have to use), and they won’t approve one that someone else has already chosen. Currently, most network numbers being distributed use the first 3 numbers (and possibly part of the 4^th), leaving only the last number for the addresses of hosts on the network. Yes, this limits you to 255 devices on your network that is connected to the Internet – but there are workarounds to this, some of which we’ll cover later.

Three classes of addresses that are in general use. The “Class” designates how much of the 32-bit number is used for the network number part of the IP address. These classes are:

II  TCP/IP Addressing Overview 1 Class A