SME Network Internet IP Addressing Strategies
By Kerry Anders
Every device that needs to communicate with the Internet needs a unique IP address. Without
this, the Internet doesn't know where to send replies to your requests. There are two different, and
functionally incompatible, IP addressing schemes:
• IPv4, which is based on a single 32-bit number that supports rather fewer
than 2^32 (4 billion) unique addresses. This version is used almost universally at the moment.
Given that every PC, server, router etc on the Internet needs a unique address, we are rapidly
running out of addresses.
• IPv6, which is intended to replace IPv4, uses a 128-bit number that can
support about 3.8 x 10^38 (340 undecillion) addresses. That is a lot of addresses, and, although
most operating systems support it, currently, it is used on less than 1% of installed equipment.
So, given that, for the moment, we are stuck with IPv4, how do we deal with the shortage
of addresses?
There are two main strategies within IPv4 to reduce the number of addresses on the Internet,
and they work together:
1. Private and Public address ranges.
IP addresses are made up of 32 bits, arranged as four "octets". Each octet can take any
value from 0 to 255, and so they are normally written as four numbers separated by periods,
e.g. 202.12.27.33
Blocks of IP addresses are allocated to the various organizations that manage the provision
of addresses to businesses and the general public.
Certain blocks are designated as "private" while all the others are "public". Private
addresses are not allowed to be connected to the internet directly, and any data bearing a
private address as source or destination will be blocked by the Internet's routers: any network
that needs to exchange data with the Internet must have at least one public IP address.
Because public addresses are in short supply, ISPs use a number of strategies to minimize
their use:
• Normally, ISPs dynamically allocate a single public IP address to each
end-user, so that each time they log in, they are given a different address from the "pool"
of addresses. Such dynamic addresses are issued by the ISP's DHCP server and will probably
"expire" after a while.
Hence, if you leave your Internet router on all the time (as most people do nowadays)
you may be disconnected when the DHCP server's address "lease" expires, and you will need
to disconnect your Internet connection and log in again. Some ISPs manage this better
than others. If you want to be sure it doesn't happen, then ask your ISP for a dedicated (static)
IP address for your Internet connection.
• Most ISPs will assign you a single IP address (or sometimes 2) at no extra
charge. If you want more that this (say you have 3 PCs at your home and want them to have unique
public IP addresses, you will have to pay extra for the other addresses.
2. NAT (Network Address Translation)
If you are limited to a single public IP address but have more than one PC on your network,
you will need to use NAT to 'map' each of the PCs onto a single address. Normally, the NAT
function is included in your Internet router.
So how does NAT work?
The TCP/IP protocol suite (as well as many other networking protocols) recognize 64K
(65,536 or 2^16) different "ports" associated with each IP address. This number is far in excess
of the number required for any one PC, so NAT creates a translation list for service that each
PC on your network requests to an Internet-facing port on your public IP address. Replies to
requests are routed either to the requesting port or to an associated port, and so the Internet
router can intercept the replies from the Internet and direct them to the correct PC. Of course,
it's not quite that simple.
| 
About BTH
User Agreement
Privacy
Site Map
Contact
News Feeds
RSS Feed
Follow @Stephen Bucaro
Basics and Certification
