11.4 Introduction to IPv6

In this video I’m going to give a very brief introduction to IPv6. Up to now we’ve only been looking at IPv4 addresses and you may remember from earlier sessions that we’re running out of them. Despite the introduction of NAT and subnetting there still aren’t enough public addresses to meet all current and future needs and this is the main reason that IPv6 was developed.

You should recall that IPv4 has address that use 32 bits, which provide over 4 billion addresses. An IPv6 address has 128 bits, which doesn’t sound that much bigger but if you try raising 2 to the power of 128 you’ll find that in decimal it produces a number that’s too big to say with ease. It’s usually referred to as ‘around 340 undecillion’. The IPv6 address space is so big that according to one well-used quotation ‘we could assign an IPv6 address to every atom on the surface of the Earth, and still have enough addresses left to do another 100+ Earths.’

An IPv6 address is written as eight groups of 4 hexadecimal numbers, each group separated by a colon. (You were introduced to hexadecimal in an earlier session.) Long strings of numbers like this are very difficult for humans to read, even when separated by colons. But there is a way of shortening them, which simply involves removing some of the zeros. This is done by a two-stage process. First, completely removing any leading zeros at the start of each block. Next, replace any consecutive zero blocks by double colons – though this can only be done once. If there is an additional consecutive zero block, this has to remain. Can you think why?

The reason is that the original address is recovered by reinserting the zeros until there are again 8 blocks of 4 hexadecimal numbers. If you were to remove more than two consecutive blocks of zeros, wewouldn’t know how many blocks to reinsert for each double colon.

Next we’ll look at the IPv6 address format. Like IPv4 it has two components: one that identifies the network – called the network prefix or ID – and one that identifies the host – called the interface ID. Note the different terminology used here: IPv4 refers to a host portion while IPv6 refers to an interface portion. The IPv6 terminology is much more correct as a network device can have a number of different interfaces, each with its own IP address. For example, think of a router with its multiple interfaces, or a laptop, which typically has an interface for its wired connection and a Wi-Fi interface. Each with its own IP address, if they are used simultaneously.

In IPv4 both address portions have variable lengths but in IPv6 these lengths are fixed, both being portions of 64 bits.

In earlier sessions you learned that the highest IPv4 network address was reserved for broadcasting to all devices on the same network. In IPv6 there is no broadcast transmission, but there are three other modes that are available. Unicast for transmissions from one interface to another – these addresses are what most of us use most of the time for sending and receiving emails or web browsing; multicast for transmissions from one interface to multiple other interfaces; and anycast for transmission from one interface to one other from a list of multiple interfaces. Each of these three address modes has a specified address structure to identify what type of address it is, but I won’t go into that here as this is just a general introduction to IPv6.

The increased address space of IPv6 gives it its main advantage over IPv4, but there are other benefits too. I’ll briefly identify them, but it’s beyond the scope of this course to explain them. IPv6 results in more efficient routing by using a hierarchical addressing structure. Its packet header structure is simpler than IPv4’s. The increased address space means there is no need for NAT in an IPv6-only network. IPv6 packets have a larger payload giving improved throughput and efficiency. IPv6 is more secure than IPv4.

It’s never been suggested that one day we will switch off IPv4 and everybody will start using IPv6. This is because there are ways of handling both IPv4 and IPv6 over the same network. This, coupled with the development of subnetting and NAT, helps to explain why the uptake of IPv6 has been gradual. When I checked this Google page in April 2019, it told me that only 27% of its users were connected over IPv6. You might like to check this out for yourself to see how it has changed since then.