Subnet Addressing



Subnet Addressing


Subnet addressing is used by system administrators in order to further subdivide an Internet
address within an organization. By using part of the host portion to identify a subnet, an
organization need only obtain one IP address from the NIC.

For example, let's pretend we are responsible for the network setup of an international
organization that has locations in the US and Europe (figure 9).



Figure 9: Subnet Mask


The US has three locations, each with a router. Europe has 2 locations with routers. We
have obtained a Class A IP Address for our network of 11.0.0.0. We now need to create
unique network numbers for each side of the routed network.

We need to decide which bits of the host address to use as part of the network number.
Although any bits of the host can be used, it is easiest to keep all host address bits on the
right side of the IP address, and all network number bits to the left side.

Let's use the highest three bits of the host address area for the subnet mask. The bits of our
address is then divided into:

network number: nnnnnnnn.ssshhhhh.hhhhhhhh.hhhhhhhh

Where n equals a bit reserved for the network number assigned from the NIC,
s equals a bit reserved for the subnet, and h equals a bit reserved for the
host address. The first bit of the subnet will equal 0 if the packet is to be routed to the US,
and 1 if it is to be routed to Europe. The remaining two bits of the subnet will be used to
differentiate the routers within the continent. For example, the routers in the US will have
the subnet mask values of 000 (subnet A), 001 (subnet B), and 010 (subnet C). Therefore,
the network numbers of the subnets located in the US are

00001011.00000000.00000000.00000000 (11.0.0.0),
00001011.00100000.00000000 (11.32.0.0), and
00001011.01000000.00000000.00000000 (11.64.0.0).

Because the subnet bits have been divided logically based on their routes, it will be easier to
intuitively determine which subnet a packet is destined for. To define the subnet mask, all
bits used for the network portion are set to 1's. All bits used for the host portion are set to 0.
The subnet mask in this example is


11111111.11100000.00000000.00000000B, or 255.224.0.0D.


The variable for setting the subnet mask is found in PROTOCOL.INI under the
subnetmask0 entry of the [TCPIP_XIF] section. Commonly, an organization will use a
full byte to use for subnet masking. This makes subnet addresses easier to read and it is
easier to administer. Following the example, a company was given the Class A address of
11.0.0.0. It decided to use the subnet mask of 255.255.0.0. One of the workstations in the US
had the IP address of 11.1.1.69. The network portion is:

11.001.001.069 = 00001011.00000001.00000001.01000101
&
255.255.000.000 = 11111111.11111111.00000000.00000000
---------------------------------------------------------
11.001.000.000 = 00001011.00000001.00000000.00000000

To find the destination address for a directed broadcast for net 11.1.0.0, you take ~ of the
subnet mask (I.E.: ~255.255.0.0 = 0.0.255.255) then bitwise OR it with the IP address, I.E.:

11.001.001.069 = 00001011.00000001.00000001.01000101
|
000.000.255.255 = 00000000.00000000.11111111.11111111
---------------------------------------------------------------
11.001.255.255 = 00001011.00000001.11111111.11111111

The following sections will detail the protocols implemented in Microsoft's TCP/IP stack. It is
recommended for detailed information on the protocols that you refer to Comer's book.



Back to Index Page.