TCP/IP network addressing concepts
An IP address, e.g. 140.106.245.102, is commonly displayed as four decimal values, separated by decimal points or “dots”. This is referred to as “dotted decimal notation”. The first decimal value, in this example 140, can be used to identify the class of addresses to which this address belongs. In addition, each class of address has an associated default subnet mask. The following table shows the class of addresses and subnet masks: First decimal value 1 – 127 128-191 192-223 224-239 240-254 Default subnet mask 255.0.0.0 255.255.0.0 255.255.255.0
Class of address A B C D E
The four decimal values, or 32 bits, that make up a TCP/IP address are used to identify the network ID, the host ID and, optionally, the subnet ID for the specific address in question. In a network that has not been subnetted, the four decimal (32 bits) values associated with the address are subdivided according to their purpose as follows: In a Class A network, the first eight bits (first decimal value) are used to identify the network ID; the remaining 24 bits (last three decimal values) are used to identify the host ID. The default subnet mask for a Class A network is 255.0.0.0. Any other subnet mask used with a Class A network would be considered a custom subnet mask and would indicate that the network has been subnetted. In a Class B network, the first sixteen bits (first and second decimal values) are used to identify the network ID; the remaining sixteen bits (last two decimal values) are used to identify the host ID. The default subnet mask for a Class B network is 255.255.0.0. Any other subnet mask used with a Class B network would be considered a custom subnet mask and would indicate that the network has been subnetted. In a Class C network, the first 24 bits (first three decimal values) are used to identify the network ID; the remaining 8 bits (last decimal value) are used to identify the host ID. The default subnet mask for a Class C network is 255.255.255.0. Any other subnet mask used with a Class C network would be considered a custom subnet mask and would indicate that the network has been subnetted. When a network is to be subnetted, the network designer determines how many subnets are necessary, considering future growth and needs, and/or how many hosts need to exist on a single subnet. When this determination is made, a custom subnet mask must be created to allow for the subnets and hosts necessary. We create the custom subnet mask by “borrowing” bits that normally would be used as the host ID and use them instead as a
subnet ID. We identify the “borrowed” bits by masking them, or making them “1s” in the subnet mask associated with the network. For example, the address 140.106.53.23, with a subnet mask of 255.255.240.0, would identify a specific address on a Class B network that has been subnetted. Remember that normally for a Class B address, the first sixteen bits (first and second decimal values) are used to identify the network ID and the remaining sixteen bits (last two decimal values) are used to identify the host ID. We know that this network has been subnetted because it is not using the default subnet mask (255.255.0.0) for this class of address. Because it is a Class B address, the first sixteen bits (two decimal values) still identify the network ID. Now, however, the last sixteen bits, normally used to identify the host ID, are used to identify both the subnet ID AND the host ID. We now must consider the binary equivalent of the third decimal value (240) in the subnet mask: 240 = 1 1 1 1 0 0 0 0 In this value, four of the bits that normally would have been used as the host ID have been masked (changed to “1s”), indicating that these four bits are now used as the subnet ID. Using four bits to identify the subnet ID, we can have 2^4 or 16 possible combinations, resulting in 14 valid subnet addresses. (We reduce the number of possible combinations by two to account for the invalid subnet IDs of all zeroes or all ones.) This also means that, instead of the normal sixteen bits to identify the host ID on a Class B network, we now have only twelve, since we used four of the sixteen as the subnet ID. Using twelve bits, we can have 2^12 or 4096 possible combinations resulting in 4094 valid host IDs. (Again, we reduce the number of possible combinations by two to account for the invalid host IDs of all zeroes or all ones.) The result is that this Class B network, when subnetted as above, results in 14 possible subnets each containing a maximum of 4094 hosts; a total of 57316 (14 x 4096) valid addresses for this network. Notice, that had we not subnetted and had used sixteen bits only for the host ID, we would have had a single network with 65,534 (2^16 - 2) valid addresses. By subnetting, we have created some invalid addresses and reduced the total number of valid, available addresses. If we consider the same network address, but use the subnet mask 255.255.252.0, we determine the results in a similar manner. Consider the decimal value 252: 252 = 1 1 1 1 1 1 0 0 In this case six bits, normally part of the host ID have been masked (changed to “1s”), indicating that they are now used as the subnet ID. This results in 62 (2^6 – 2) valid subnets. There are now ten bits (the usual 16, less the 6 we “borrowed” for the subnet ID) remaining to identify the host ID. That allows for 1022 (2^10 - 2) host IDs. Using 62 subnets, each with a maximum of 1022 hosts, we have 63364 valid addresses. Again, note how the number of host addresses (65,534) normally available on a Class B network has been reduced by subnetting.