1. IPv6 Adapted from Forouzan TCP/IP
Further reading:
RFC IP Version 6 Addressing Architecture

2. IPv6 Why IPv6? IPv6 ADDRESSES IPv6 PACKET FORMAT ICMPv6
TRANSITION - IPv4 TO IPv6

3. Need for IPv6 Address space exhaustion
- for all RFCs
Address space exhaustion
Two level (classfull) addressing is wasteful
Network addresses are used even if not connected to Internet – dynamic scheme needed
Growth of the Internet;
New areas – e.g. POS (point of sale, online stores), cable TV receivers
Single address per host restrictive (e.g. cannot use ‘address’ to signify scope of routing, multicast, etc.)
Requirements for new types of service
Real time streaming multimedia - QoS
Retail/financial - security

4. IPv6 and IPv4 compared (1) 1995 v.s. 1975
IPv6 header (40 bytes) only twice the size of IPv4 header
Only version number has the same position and meaning as in IPv4
Simplified header, removed: header length, type of service, identification, flags, fragment offset, header checksum
Datagram length replaced by payload length
Protocol type replaced by ‘next header’

5. IPv6 and IPv4 compared (2) Time to Live (TTL) replaced by hop limit
Added: Priority and flow label
All fixed size fields
No optional fields. Replaced by extension headers
8-bit hop limit = 255 hops max (Limits looping)
Next Header = 6 (TCP), 17 (UDP)

6. (IPv6 Features and Advantages)
IPv6 and IPv4 compared (3)
(IPv6 Features and Advantages)
Bigger address space – 128 bits, 16 bytes
Efficient and Extensible IP datagram
Efficient Route Computation and Aggregation
Improved Host and Router Discovery
New Stateless and Stateful Address Autoconfiguration
Multiple addresses per interface
Dynamic address configuration
Better header format – simple main header
New Options – in optional headers
Allows future expansion
Support for resource allocation, QoS
Security – encryption, authentication

7. IPv6 address
2 bytes each

8. Abbreviated address with consecutive zeros
A sequence of zeros of any length can be abbreviated to a double colon ::
But this can only be done once. e.g. FDEC:0:0:0:0:BBFF:0:FFFF  FDEC::BBFF:0:FFFF
We can count four groups, so the :: must represent four zeros. The second run of zeros must be included. If we abbrevaited to:
FDEC::BBFF::FFFF, we won’t know how many zeros are contained in each group (4/1 or 3/2 or 2/3 or 1/4)
Allows classless addresses (CIDR)

9. Address structure
Categories – unicast, multicast (to a group), anycast (nearest one of a group)
Types (see next slide):
Provider based; Geographic based; Link local; Site local; Network Service Access Point (NSAP) for OSI addressing schemes; IPX; etc.

10. Address Allocation Start of address (type prefix) % Fraction
0.39
1/256
Reserved
Unassigned
0.781
1/128
Reserved for NSAP allocation (OSI addressing)
IPX allocation
010
12.5
1/8
Provider based unicast
100
Geographic based unicast
0.1
1/1024
Local use (link) FE 10
Local use (site) FE 11
Multicast FF
How are the percentages worked out?
is one permutation out of a possible 256, 1/256 = .0039
010 is one of a possible 8, is 12.5%

11. Provider-based (ISP) address
Recommended: 16,24,32,48 bits for provider, subscriber, subnet and node identifiers. Hierarchical addressing – to know the subscriber, you need to know the registry, provider as well.
Authority with whom the provider is registered

12. Unspecified address Loopback
Equivalent to the IPv4 unspecified address of
Typically used in the source field of a datagram sent by a device seeking to have its IP address configured.
Loopback

13. IPv4 IPv6 – Mapped/Compatible
For Hybrid dual-stack IPv6/IPv4 implementations
(support translations between the stacks)
IPv4 addresses that have been mapped into the IPv6 address space, and are used for devices that are only IPv4-capable.
v4 address Mapped to v6 address space: v4  (v6…v6)  v4
Compatible: Communication between v6 nodes in a sea of v4: v6  (v4…V4)  v6 (96 0’s + IPv4 address)
Mapped: destination is v4, source may be v6 or v4, via a v6 island. v4 or v6 computer  (v6…v6)  v4 (80 0’s ’s + 32-bit IPv4 address)
Note: For both types – compatible and mapped, the checksum is the same whether 4 bytes or 16 bytes are used. The extra 0’s and 16 1’s do not make a difference.

14. IPv4 IPv6 – Mapped/Compatible
For Hybrid dual-stack IPv6/IPv4 implementations
(support translations between the stacks)
IPv4-compatible IPv6 address, used for devices that can understand both IPv4 and IPv6;
v6 address Compatible with v4: v6  (v4…v4)  v6 or v4
Compatible: Communication between v6 nodes in a sea of v4: v6  (v4…V4)  v6 (96 0’s + IPv4 address)
Mapped: destination is v4, source may be v6 or v4, via a v6 island. v4 or v6 computer  (v6…v6)  v4 (80 0’s ’s + 32-bit IPv4 address)
Note: For both types – compatible and mapped, the checksum is the same whether 4 bytes or 16 bytes are used. The extra 0’s and 16 1’s do not make a difference.

15. Link local address FE 10….. (to isolate a LAN)
Not forwarded by any router
Site local address FE 11……
site wishes to stay isolated from the Internet
Forwarded only within a ‘site’ (user defined, could be autonomous system).
Not forwarded by border router

16. Multicast address
Permanent/transient – whether the multicast group exists forever or is a temporary thing.
Scope: how far does the multicast group extend?
Predefined groups: all nodes (1), all routers (2), all DHCP servers (1:0)

17. Example of a permanent multicast address FF 0x
For example 43 = (user assigned) group of all NTP servers (Network Time Protocol)
FF01::43 on this node
FF02::43 on this link
FF05::43 on this site
FF08::43 in this organisation
FF0E::43 global
Other examples of groups:
All name servers
All routers
FF = multicast, 0 = permanent, x specifies scope.

18. More Examples – link local
Scope: Link level
Permanent
Multicast group 67 (0x 43) is FF02::43
Transient
Multicast group 317 (0x 13D) is FF12::13D

19. IPv6 datagram

20. Format of an IPv6 datagram
Each row represents 4 bytes
What is the length of the standard header?

21. IPv6 header fields Version: The number 6 encoded (bit sequence 0110).
Priority (Traffic class): The packet priority (4 bits). Priority values subdivide into ranges
-traffic where the source provides congestion control,
-non-congestion control traffic.
Flow label: Used for QoS management (20 bits). Originally created for giving real time applications special service (currently unused).
Payload length: The size of the payload in octets (16 bits).
When cleared to zero, the option is a “Jumbo Payload" (calculated hop-by-hop).
Next header: Specifies the next encapsulated protocol. The values are compatible with those specified for the IPv4 protocol field (8 bits).
Hop limit: Replaces the TTL (Time to Live) field of IPv4 (8 bits).
Source and destination addresses: 128 bits each.

22. Priority (4 bits) Priority – 0 to 15 allowed; 0-7 ‘ordinary’
0 lowest, unspecified
1 background (e.g. news)
2 unattended data traffic e.g.
3, 5 reserved
4 attended bulk data – FTP, HTTP
6 Interactive traffic e.g. Telnet
7 Control traffic e.g. SNMP, RIP, OSPF
8 to 15: Delay sensitive e.g. voice, video

23. Flow labels (24 bits)
A flow = a sequence of packets defined by source & destination addresses and flow label.
Special handling (QoS) can be requested – optimum delay or throughput
A flow is usually from a single TCP connection, sometimes more (e.g. FTP – two connections 20,21)
A single application may generate several flows (e.g. audio, video)

24. Routers and flows A router has to decide on:
Path
Precedence
Resource allocation
Discard requirements - buffering
Accounting & Security
Negotiated on a per flow basis
Packets do not have to carry all this information – flow label represents this info.

25. Rules for use of flows
Hosts/routers which do not implement flows – set to 0 if initiating, ignore and leave unchanged otherwise.
All packets originating with a given flow label must have the same destination address, hop-by-hop options and routing headers (if used).
The router maintains information about flows in a hash table.
Flow labels are assigned by source at random; flow labels from a given host must be unique.

26. Extension header format

27. Next Header Codes Hop-by-Hop For all routers e.g. ‘jumbogram’ 2 ICMP
Hop-by-Hop
For all routers e.g. ‘jumbogram’ 2
ICMP
Defines the protocol at the next level 6
TCP 17
UDP 43
Source routing
Strict and loose permitted 44
Fragmentation
As for IPv4, but done by the source 50
Encrypted security payload
The payload is encrypted - transport or tunnel mode. 51
Authentication
Identifies the sender and guarantees message integrity 59
Null (no next) 60
Dest. option
Processed by the destination only

28. IPv6 Extension Headers
Routing type = 0/1 strict/loose routing

29. Hop-by-hop option header
Pad-1: for alignment – some options start at a specific bit position in the 32 bit word. It consists of’
Pad-N: allows variable no. of bytes (>=2) as needed length (1 byte) + zero or more bytes of padding.
Jumbo Payload: says it is a larger than normal (65535 bytes) packet. Up to 232
Defined options

30. Note – this is only for the specific link
The Pad1 option is used to insert one octet of padding into the Options area of a header.
The PadN option inserts two or more octets of padding into the Options area of a header. For N octets of padding, the Opt Data Len field contains the value N-2, and the Option Data consists of N-2 zero-valued octets.
Pad1 and pad-N for alignment (to 32 bit word)
The jumbo payload must start at a 2 byte boundary from the start of the extension header (4n + 2)
Note – this is only for the specific link

31. Source routing
The destination address and the list of intermediate addresses change from router to router
Source routing may be used to select a specific network or ISP on the basis of cost, current location etc.

32. Source routing example
Notice that the destination address changes from hop to hop, as well as the list of address entries in the header..

33. Fragmentation done by host, NOT routers
MTU discovery needed to determine a suitable packet size across the whole route (default 576 bytes).
Method, parameters and flags are the same.

34. Authentication (plus integrity check)
Security parameter index defines the algorithm used.
Authentication and integrity check are provided by this process.

35. Encrypted security payload
(ESP)
Index defines the algorithm. Two modes:
Transport mode: Only the TCP/UDP payload is encrypted
Tunnel mode: Entire IP packet is encrypted and encapsulated in a new IP packet with the ESP header
Provides confidentiality and guards against eavesdropping.
Security parameter index defines the algorithm used.

36. Transport mode encryption
Tunnel-mode encryption

37. Transition v4 to v6 Not all routers can be upgraded simultaneous
no “flag days”
How will the network operate with mixed IPv4 and IPv6 routers?
Proposed approaches:
Dual Stack: hosts with dual stack (v6, v4) can “translate” between formats. A few IPv4 systems in a majority of IPv6 nodes;
Tunneling: IPv6 carried as payload in IPv4 datagram among IPv4 routers

38. Dual stack
Which version to use for a particular destination? Ask DNS.

39. Dual Stack Routers A B C D E F IPv6 IPv6 IPv4 IPv4 IPv6 IPv6 A-to-B:
Flow: X
Src: A
Dest: F
data
Src:A
Dest: F
data
Src:A
Dest: F
data
Flow: ??
Src: A
Dest: F
data
A-to-B:
IPv6
B-to-C:
IPv4
D-to-E:
IPv4
E-to-F:
IPv6

40. Tunneling – IPv6 packets in a sea of IPv4 routers.B E F
IPv6
tunnel
Logical view:
Physical view: C D
IPv4
Flow: X
Src: A
Dest: F
data
Src:B
Dest: E
A-to-B:
E-to-F:
B-to-C:
IPv6 inside
D-to-E:
Automatic tunneling compatible IPv6 (destination) address is used
Configured tunneling: The border v4 routers use their own IPv4 addresses to encapsulate the v6 packet