IP Version 6 Next generation IP Prof. P Venkataram ECE Dept. IISc

Organization of the talk Why IPv6 ? Address problem IPV6 Address formats and types IPV6 Headers IPV6 to IPV4 comparision IPV6 Features IPV6 transition

Why Ipv6 ? Problem with IP v4 Limited address space, Inflexible address allocation Current solutions Use Network Address Transalation (NAT) to solve limited address space problem Use CIDR (Classless Inter Domain Routing) to solve address allocation problem Problem with NAT NAT do not work for many protocols(Multimedia, Internet gaming) IPsec not possible Bottleneck in high speed networking

Address Problem How many address are supported in IPV4 IPV4 is a 32 bit address, supports 2 32 = 4 Billion addresses Class A supports 2 24 = 15 Million nodes Class B supports 2 16 = 64 thousand nodes Class C supports 2 8 = 256 nodes Is this enough ?? Indian population is over 1 Billion If 10% of India uses internet (for voice, data OR anything) India alone needs 100 Million IP nodes Problem is elevated if each person is served by more than one computer What about rest of world !! What will happen by 2010 ??

IPV6 address Add more bits in the address, as compared to IPV4 IPV6 address is 128 bit = 3.4 x addresses !! IPV6 address supports Multiple interface per hosts Multiple address per interface Unicast, Multicast and Anycast address Provider based, site-local and link-local address hierarchy

Address Format Colon-Hex notion x: x: x: x: x: x: x:x Where x is a 16 bits hexadecimal field 2001: 0000: 1234: 0000: 0000: c1c0: ABCD: 0876 Case insensitive 2001: 0000: 1234: 0000: 0000: c1c0: ABCD: 0876 Leading zeros in a field are optional: Eg 2001: 0000: 1234: 0000: 0000: c1c0: abcd: : 0: 1234: 0: 0: c1c0: abcd: 876 Successive fields of 0 are represented as ::, but only once in an address: 2001: 0: 1234:: C1C0: ABCD: 876 Not valid: 2001:: 1234:: C1C0: ABCD: 876 Other examples: FF02: 0: 0: 0: 0: 0: 0: 1 => FF02:: 1 0: 0: 0: 0: 0: 0: 0: 1 => :: 1 0: 0: 0: 0: 0: 0: 0: 0 => ::

Address Types - Unicast Unspecified 0: 0: 0: 0: 0: 0: 0: 0 or :: Like in Ipv4 Loopback 0: 0: 0: 0: 0: 0: 0: 1 or :: 1 Like in IPv4 Scoped addresses Link-local - not forwarded outside the link new in IPv6 FE80: 0: 0: 0: Site-local – not forwarded outside the site FEC0: 0: 0: : Subnet id = 16 bits = 64K subnets Aggregatable Global 128 bits as the total 48 bits prefix to the site 16 bits for the subnets in the site 64 bits for host part

IPV6 unicast address Link-local Site-local Aggregatable Global 10 bitsN bits118 - N bits Interface ID 10 bitsN bitsM bits118 - N bits Subnet IDInterface ID 48 bits16 bits64 bits Site prefixsubnetHost part

IPV6 Multicast Address First 8 bits are always 1 Flag 0000 permanent (well known) multicast address 0001 Transient multicast address Scope defines the scope of the packet Node local(1), Link Local(2), Site Local(5), Organizational Local(8), Global(E) Group ID 12 bit multicast group identifier 8 bits4 bits 112 bits 1111 FlagScopeGroup ID

IPV6 Address Allocation

IPv6 Header compared to IPv4 Header

IPv4 Vs IPv6 Source and destination addresses 32 bits (4 bytes)128 bits (16 bytes) IPsec support OptionalStandard None in headerIncluded in header Fragmentation By both routers and sending host Only by sending host Header checksum IncludedNot included Header optional data Included Moved to extension headers Identification of packet flow for QoS handling by routers IPv4IPv6

Extension Headers IPv6 header Hop- by- Hop Options header Destination Options header Routing header Fragment header Authentication header Encapsulating Security Payload header Destination Options header Upper- layer header

IPV6 Features Larger Address Space From 32 bits to 128 bits addresses enables: Global reachability No hidden networks, hosts All hosts can be reachable and be "servers" End- to- end security can be used Flexibility Multiple levels of hierarchy in the address space Autoconfiguration Use of 64 bits for link- layer address encapsulation with warranty of Uniqueness

IPV6 Features "Plug and play" By autoconfiguration Aggregation Multiple prefixes for the same site enables multihoming without cutting holes in the aggregation Multihoming Renumbering By using autoconfiguration and multiple prefixes, renumbering becomes doable Built-in Mobility Support Built-in Security Support

IPv6 Transition When moving to another technology, the transition has to be discussed and is generally very important. Often it is where most of the money is put Many new technologies didn’t succeed because of the lack of transition scenarios/ tools IPv6 was designed, at the beginning, with transition in mind Transition Strategy For end- systems, there is: Dual stack approach For network integration, there are: Tunnels IPv6-only to Ipv4-only translation

Dual Stack Host Node has both IPv4 and IPv6 stacks and addresses IPv6- aware application asks for both IPv4 and IPv6 addresses of destination DNS resolver returns IPv6, IPv4 or both addresses to application IPv6/ IPv4 applications choose the address and then can communicate With IPv4 nodes using IPv4 Or with IPv6 nodes using IPv6

Tunnelling IPv6 in IPv4 IPv6 encapsulated in IPv4 Many topologies possible Router to router, Host to router, Host to host The tunnel endpoints take care of the encapsulation. This process is “transparent” for the intermediate nodes

Conclusion

References RFC Internet Protocol, Version 6 (IPv6) Specification RFC Neighbor Discovery for IP Version 6 (IPv6 ) RFC Key and Sequence Number Extensions to GRE RFC IPv6 Tunnel Broker RFC Transmission of IPv6 over IPv4 Domains without Explicit Tunnels RFC IPv6 Testing Address Allocation