Historical summary and
reasons behind its development
In the early
nineties, the Internet Engineering Task Force (IETF)
started developing the successor to IPv4. Various
parallel efforts were being made to resolve the
problem that was expected with regard to space
constraints and the functionality of the new
version. That is how the IETF started up the Next
Generation Internet Protocol (IPng) in 1993, which
was aimed at examining different proposals and
recommendations on procedures to be considered in
the future.
Once these working
teams were set up, they recommended the creation of
IPv6 at a meeting held by IETF in Toronto in 1994.
Their recommendation was embodied in Request for
Comments (RFC) 1752, and the proposed document
was entitled “The Recommendation for the IP Next
Generation Protocol,” which is when it is considered
that the new protocol was born. The working group
then started studying the life expectancy of the
remaining IPv4 addresses and calculated that current
IPv4 addresses would run out by 2005-2011.
During this period,
supplements were added to the standard to allow
industry to participate and enable subsequent
developments; this was presented in a standard
called “Internet Protocol Version 6 (IPv6)” in 1995,
and almost immediately afterwards, it continued to
be revised and was presented in RFC 2460.
In the framework of
the new adoption of the technology, some companies
have started to experiment on IPv6, among which the
following are noteworthy: Microsoft launched a new
version of Windows with support for IPv6, Cisco
Systems introduced the IPv6 support for its Cisco
IOS routers (2001) and the Google search engine with
its site in ipv6.google.com.
What does IPv6 offer?
At first, IPv6
should be viewed as an evolution of IPv4, ever since
companies were allowed to participate, the goal of
keeping the operating and software systems
up-to-date was achieved. At present, there are
mechanisms available for a step-by-step transition
for switching to the current infrastructure of IPv6,
without jeopardizing the current network system. It
is now important to examine the proposal on the
basis of its changes: larger address space,
autoconfiguration, simplification of header format.
These features are specified below:
·
Larger address space
The address format
is enlarged from 32 bits to 128 bits. This provides
an IP address for each grain of sand on the planet.
Furthermore, it also makes it possible to
hierarchically structure the address space promoting
overall routing optimization.
·
Autoconfiguration
Probably the most
intriguing feature of IPv6 is the autoconfiguration
mechanism, which is designed as follows: when a
starting device in IPv6 requests the network prefix,
it can receive one or more network prefixes from an
IPv6 router present in its connection. The
information of the prefix can be automatically
configured if MAC or fixed addresses are used. In
the IPv4 world, we have to assign a unique IP
address to each device either by manual
configuration or by DHCP. With autoconfiguration,
however, network administrators have a much easier
life and time spent on IP network maintenance is
saved. Furthermore, when we start imagining the
number of devices with an IP address that our
households can have in the future, this function
becomes indispensable.
·
Simplifying the header format
The IPv6 header is
much simpler than the IPv4 header and it has a fixed
length of 40 bytes. This allows for greater
processing speed. Basically, it involves 16 bytes
for the source and 16 bytes for the destination
address and only 8 bytes for the header’s general
information.
·
Extended support for new options and extensions
IPv4 integrates
options in the base of the header, whereas IPv6
carries options in the extension headers and options
are only inserted in the base headers when required.
Once again, this substantially speeds up packet
processing. The base of the specification describes
a set of six extension headers, including routing
headers, IPv6 mobility, service quality and
security.
IPv6 addresses are
comprised of 16 bytes. The first bits identify the
type of address, just as in IPv4. There are many
classes of addresses, however, but not all have the
same range allocation, and most of them are reserved
for future use. In addition, a specific range for
IPv4 addresses has been planned and thus any IPv4
address can be included in an IPv6 packet.
One part of the
address space was reserved for geographical
distribution, much like what is currently done with
CIDR. Another part was reserved for distributing
addresses by provider. The possibility of having
Internet evolve toward a network that interconnects
the networks of the large providers throughout the
world, where geographical location would be
secondary, has been envisaged. For this, a
hierarchical structure of addresses with various
levels was planned. For the multicast addresses, a
specific range was planned and in the format of
these addresses, a 4-bit field was reserved, making
it possible to specify the range that the broadcast
intends to have.
No specific address
has been planned for broadcast, as this is
considered a specific case of multicast. In
addition to unicast, multicast and broadcast
transmission, anycast transmission can be made,
whereby a packet is sent to any group member,
without caring or specifying which. This makes it
possible, for example, to gain access to a
multihomed server, achieving a load balance between
the various interfaces or by the one that is closest
to the petitioner. It also facilitates redundant
configurations where a given service can be
delivered by more than one server. A range of
addresses of local significance was also considered,
equivalent to private addresses, for cases in which,
for security reasons, one wishes to be completely
isolated from the outside. The notation of IPv6
addresses is as follows: they are written in eight
clusters of four hexadecimal digits, separated by
colons.
An IPv6 address is
128-bits long and is comprised of eight fields of 16
bits, each united by a colon. Each field must
contain a hexadecimal number, in contrast to the
decimal point notation of the IPv4 addresses. In the
following figure, the “X”s represent hexadecimal
numbers.
MSc Carlos Peña
Profesor del curso
Comisión Nacional de Telecomunicaciones de Venezuela
(CONATEL) |