Internet Protocol version 6 (IPv6)

Internet Protocol version 6 (IPv6), started standardization as a replacement of the current IP version 4 (IPv4), described in RFC 791 (Standard), due to the depletion of the limited number of IPv4 addresses foreseen already in the 1990's. Up until present time, various techniques such as Classless Inter-Domain Routing (CIDR), Network Address Translation (NAT), and Multi Protocol Label switching (MPLS) have managed to delay this depletion. The IETF Internet next Generation (IPNG) working group developed IPv6, RFC 2460 (Draft Standard).

The current Internet Protocol IPv4, supports up to 4 billion addresses with 32-bit address space. While 4 billion is a lot bigger than the currently estimated 2.5 billion addresses in use by several hundred million Internet users, in practice IPv4 supports a much lower number. That is because addresses are not used efficiently. They are allocated in regional blocks, and there is an over supply in some areas of the world and other areas (e.g., Asia, Europe and Latin America) are close to running out of addresses. At the current rate of 60% efficiency, IP addresses will run out some time in the future.IPv6 128-bit address format allows for 340,232,366,920,938,463,374,607,431,768,211,456 IP addresses (340 duodecillion), enough to award one to every grain of sand on earth. Figure 1 depicts the IPv6 header format.

Figure 1 – IPv6 Header Format

Besides a 128-bit wide address range, the TCP-UDP/IPv6 protocol suite provides additional features such as mandatory security and mobility, ease of administration and auto-configuration features, built-in QoS, and more scaleable routing and robustness to mention a few. Many of these have been retrofitted in IPv4 with various limitations and decreased functionality.

Wireless will have the greatest impact on IP. The forthcoming 3G will make much greater use of IP than the previous generations of cellular radio. Until now, IP has been used as an add-on to cellular networks, in a not too distant future, cellular networks will be data oriented, as voice will be treated as another IP session within the network. The development of new radio protocols such as 802.11B (Wireless Ethernet) plus new wired serial interfaces such as IEEE 1394 (Firewire) will provide the opportunity for consumer products to require an IP address to connect to the net. 1.1. IPv6 Addressing

The IPv6 addressing architecture is described in RFC 2373. The advantage of the IPv6 addressing architecture over the IPv4 one is mainly the length of the address. While IPv4 32-bit addresses can be divided into two or three variable parts (the network identifier, the node identifier and sometimes the subnet identifier); the IPv6 128-bit addresses can support different fields within the address.

IPv6 Address Representation

There are three conventional forms for representing IPv6 addresses as text strings. The preferred form is x:x:x:x:x:x:x:x, where the 'x's are the hexadecimal values of the eight 16-bit pieces of the address, however certain styles of IPv6 addresses may contain long strings of zero bits that can be represented by "::". The third alternative is a mixed environment of IPv4 and IPv6 such as x: x:x:x:x:x:d.d.d.d, where the 'x's are the hexadecimal values of the six high-order 16-bit pieces of the address, and the 'd's are the decimal values of the four low-order 8-bit pieces of the address (standard IPv4 representation).

IPv6 Address Types

There are three types of addresses in IPv6 (unicast, anycast and multicast) and all of them are assigned to interfaces, not to nodes:

Unicast addresses specify a single IPv6 interface. A node can have more than one IPv6 network interface. Unicast addresses can be viewed as 128-bit field that identifies one particular interface. However, the data in the address field can be parsed out into smaller pieces of information, although all that information when put together will result in a 128-bit field that identifies a node’s interface.

Anycast addresses are IPv6 addresses that are assigned to one or more network interfaces (typically belonging to different nodes), with the property that a packet sent to an anycast address is routed to the "nearest" interface having that address, according to the routing protocols' measure of distance. Multiple nodes may be sharing the same anycast address, like a multicast address. However only one of those nodes can expect to receive a datagram sent to the anycast address.

Multicast addresses, like broadcast addresses, are used in local networks like Ethernet, where all nodes can sense all transmissions on wire. However, IP multicast is more complicated because all packets are not forwarded to all nodes in the network; instead, the packets are only forwarded to members of the multicast group. When a node subscribes to a multicast address, it announces that it wants to become a member and any local router will subscribe on behalf of that node.

Oscar Avellaneda
Chair
Working Group on Technology
PCC.I
e.mail: [email protected]