Extremely Low prices on Ink and Toner Cartridges

TCP Addressing | Tech Info | TCP/IP |

TCP addressing

IP addresses are classified as A, B, C, D or E. They are configured in a four byte dotted decimal format such as

aaa.bbb.ccc.ddd

There are always 4 sets of numbers and each number ranges between 0 and 255 (with some exceptions). An IP address has an imaginary line separating the full address into 2 halves, the network address and the host address. Doing this allows an address to have one of more networks, each of which can have 1 or more hosts (computers connected to the network).

The complete address is used in such a way that knowing the class of the address tells you how to split the address into the 2 components "network" and "host". Class A addresses use the first number (byte), shown as "aaa" above, for the network addressing leaving the other 3 bytes (bbb.ccc.ddd) to be used in identifying the host.

Class B addresses use the first 2 bytes (aaa.bbb) for network identification and the last 2 bytes for host identification.

Class C addresses use the first 3 bytes for the network address and the last byte for host addressing.

 

Class A allows for 126 Networks and 16,777,214 hosts , and the 1st byte is 1 - 126
Class B allows for 16, 386 networks and 65,534 hosts , and the 1st byte is 128 - 191
Class C allows for 2,097,162 networks and 254 hosts , and the 1st byte is 192 - 223

a special address is the address 127 which is reserved as a Loopback address

The leftmost number determines the type of address A through E.

Looking at the first byte after converting it to binary, you will see that the class A addresses all have a 0 for the first bit.

Class B addresses all start with 10, class C with 110, class D with 1110 and class E with 1111.

The address is sometimes described using the "N" (network) and "H" (host) letters to show which part of the address is the network as which is the host.

For example a class "A" address would be N.H.H.H, class "B" would be N.N.H.H , and class "C" would be N.N.N.H.

The above designation makes more sense when you also look at how a mask is applied. A mask is used to further subdivide a set of available host addresses into a split of more networks and hosts. A mask is in the same form as the IP address and has at a minimum the following format:

Class "A" is 255.0.0.0
Class "B" is 255.255.0.0
Class "C" is 255.255.255.0

In binary 255 is 11111111 ( eight ones). So looking at the class mask in binary and running all four parts together, you see ones to the left and zeros to the right.

So for a class "A", you get 11111111.00000000.00000000.00000000 . This mask is one that has all the available hosts under one network (since all the values to the right of the ones are zeros).

To subdivide the address you come up with masks by basically replacing the zeros with ones making sure that you work from left to right and never inserting any zeros between the ones. (It is not necessary to NOT alternate zeros and ones but it sure makes it easier to understand and use).

All the time keep in mind that you have to know in advance address class. Since you know the addresses have ones up to a certain point depending on the address you only have to worry about working with the numbers to the right of the point where the address boundary is for that class.

For example a Class "B" address would have for a mask

11111111.11111111.00000000.00000000

now by replacing zeros with ones going from the left to the right you can come up with different masks.

11111111.11111111.11111111.00000000 takes the address and gives you a split where the ones in the 3rd byte (underlined) represent the amount of subnets you are allocating and the equivalent value of the zeros (if you were to change them to ones and convert them to decimal then subtract 1 ) would tell you how many hosts can be supported for every one of the individual networks that you can have.

In the above example, you know its a class "B" address so you ignore the first 2 bytes (the 16 ones) and look at the remaining 2 bytes (8 ones and 8 zeros).

You know that you will split the address into as many networks as the number represented by the ones located after the split. After the split in this case are eight ones (11111111) which converts to 255 in decimal so you have 255 minus 1 which is 254 possible networks. Now there remaining 0 zeroes can be used to figure out how many hosts you can have on each subnet. You convert all the zeros to ones and convert that number to decimal and subtract one. Since there are eight zeros you have 11111111 which is 255, subtract one and you have 254 possible hosts.

Another example, the mask

11111111.11111111.11110000.00000000

Ignore the first 2 bytes (the ones in italics),

The number of networks available is binary "1111" (underlined) which when converted to decimal is 15, minus 1, makes it 14. So with the above mask you can have 14 subnetworks.

Since there are 12 zeros, you change them to ones, and convert the 12 ones (binary) to decimal giving 4095, then subtract 1 resulting in 4094. So the mask can give you 4094 hosts for each subnetwork.

.back to the TCP/IP page

© 1997,98,99,2000 www.techadvice.com All rights reserved

---