The Future Of Tcp/ip (ipv6)

The Future of TCP/IP (IPv6)  

Chapter 33 Evolution of TCP/IP intertwined with evolution of the global Internet   



Internet is largest installed internet Funding comes from organizations that are Internet users Most researchers use Internet daily

Chapter purpose is to consider ongoing evolution of TCP/IP

© MMII JW Ryder

CS 428 Computer Networking

1

Why Change? 

New computer and communication technologies 



New applications 



New technologies = new possibilities and needs New ways to use Internet means new protocols needed

Increases in size and load 

Massive growth means old ways strained

© MMII JW Ryder

CS 428 Computer Networking

2

Motivation for Changing IPv4   

New countries with differing administrative policies IPv4 same for about 20 years Since IPv4 designed     

Enhanced processor performance Memory size increased Network bandwidth for Internet backbone increased New LAN technologies Number of hosts on Internet risen to over 56 million

© MMII JW Ryder

CS 428 Computer Networking

3

Road to New Version of IP 

Several suggested designs 

 

Make IP more sophisticated at expense of increased complexity and processing overhead Use a modification of OSI CLNS protocol Retain most of ideas in IP but make simple extensions to accommodate larger addresses  

© MMII JW Ryder

Simple IP – (SIP) Still include new ideas from other suggested protocols CS 428 Computer Networking

4

Features of IPv6 



Despite many conceptual similarities IPv6 changes most protocol details Completely revises datagram format 

 

Replace IPv4 variable length fields with a series of fixed format headers

Still supports connectionless delivery Allows sender to choose datagram size but requires sender to specify maximum hops

© MMII JW Ryder

CS 428 Computer Networking

5

Features of IPv6 Includes facilities for fragmentation and source routing  Main changes introduced are 1. Larger Addresses: IPv6 quadruples the size from 32 bits to 128 bits 2. Extended Address Hierarchy: Creates ability to have additional address levels on an internet 

 

IPv4 Addresses – 2 levels, Network and Host IPv6 Addresses – Can define a hierarchy of ISPs as well as hierarchy within a site

© MMII JW Ryder

CS 428 Computer Networking

6

Features of IPv6 3. Flexible Header Format: Datagram format entirely different 

Defines a fixed size (40 octets) header with optional extended headers

4. Improved Options: 

Has same options as IPv4 plus some new ones

5. Provision for Protocol Extension: 

Move away from protocol that fully specifies all details to one that permits additional features

© MMII JW Ryder

CS 428 Computer Networking

7

Features of IPv6 6. Support for Autoconfiguration and Renumbering: 



Allows computers on an isolated network to assign themselves addresses and begin communicating without depending on a router or manual configuration Facility to permit a manager to renumber networks dynamically

© MMII JW Ryder

CS 428 Computer Networking

8

Features of IPv6 7. Support for Resource Allocation: 

Two facilities for pre-allocation of network resources  

© MMII JW Ryder

a Flow abstraction a Differentiated Services specification

CS 428 Computer Networking

9

IPv6 Address Space  

 

How big is 2128 ? So large that everyone on earth will have enough addresses to have their own internets with as many addresses as the current Internet has So large that there would be 1024 internet addresses per each square meter on earth So large that the address space is greater than 3.4 * 1038 

If addresses are assigned at the rate of 1,000,000 every microsecond (1/1,000,000th of a second), it would take more than 1020 years to assign all possible addresses

© MMII JW Ryder

CS 428 Computer Networking

10

IPv6 Colon Hexadecimal Notation 

128 bit number expressed as dotted decimal



104.230.140.100.255.255.255.255.0.0.17.128.150.10.255.255 becomes 68E6:8C64:FFFF:FFFF:0:1180:96A:FFFF





Hex notation allows zero compression 

 

A string of repeated zeros is replaced with a pair of colons FF05:0:0:0:0:0:0:B3 becomes FF05::B3 Can be applied only once in any address

© MMII JW Ryder

CS 428 Computer Networking

11

Zero Suppression   



0:0:0:0:0:0:128.10.2.1 becomes ::128.10.2.1 Looks quite similar to IPv4 12AB::CD30:0:0:0:0/60 says use first 60 bits and becomes 12AB00000000CD3

© MMII JW Ryder

CS 428 Computer Networking

12

Basic IPv6 Address Types 



Unicast – Destination address specifies a single computer. Route datagram along shortest path. Anycast – Destination is a set of computers, possibly at different locations, that all share a single address. Route datagram along shortest path and deliver to exactly one member of the group (i.e. closest member)

© MMII JW Ryder

CS 428 Computer Networking

13

Basic IPv6 Address Types 

Multicast - Destination is a set of computers, possibly at different locations. One copy of the datagram will be delivered to each member of the group using hardware multicast or broadcast if viable.

© MMII JW Ryder

CS 428 Computer Networking

14

Encoding IPv4 Addresses in IPv6 80 zero bits

16 bits

32 bits

0000 . . . . . . . . . . . . . . . . . . . . . . . . 0000

0000

IPv4 Address

0000 . . . . . . . . . . .. . . . . . . . . . . . . . 0000

FFFF

IPv4 Address

• 16-bit field contains 0000 if node also has a conventional IPv6 address and FFFF if it does not. RESERVED © MMII JW Ryder

DATAGRAM IDENTIFICATION CS 428 Computer Networking

15

General Form of IPv6 Datagram Optional Base

Extension

Header

Header 1

 

Extension Header N

Data

40 octets

© MMII JW Ryder

CS 428 Computer Networking

16

IPv6 Base Header Format   





See Base Header figure Alignment changed from 32 bit to 64 bit multiples Header length eliminated – Replaced with PAYLOAD LENGTH field Size of source and destination addresses changed to 16 octets Fragmentation information moved out of fixed fields in base header to extension header

© MMII JW Ryder

CS 428 Computer Networking

17

IPv6 Base Header Format 





TIME-TO-LIVE field changed to HOP LIMIT SERVICE-TYPE field renamed to TRAFFIC CLASS and extended with FLOW LABEL field PROTOCOL field replaced with a field that specifies type of next header

© MMII JW Ryder

CS 428 Computer Networking

18

Base Header Format 0

4

VERS

12

16

TRAFFIC CLASS

PAYLOAD LENGTH

24

31

FLOW LABEL NEXT HEADER

HOP LIMIT

SOURCE ADDRESS

DESTINATION ADDRESS

Base Header Size: 4 + 4 + 16 + 16 = 40 Octets © MMII JW Ryder

CS 428 Computer Networking

19

Base Header Format 

PAYLOAD LENGTH is length of all extension headers plus data 



i.e. Total length – 40 octets (Base Header)

IPv6 datagram can contain up to 64K octets of data

© MMII JW Ryder

CS 428 Computer Networking

20

Traffic Class   

IPv4 SERVICE CLASS renamed to TRAFFIC CLASS New IPv6 mechanism allows for resource reservation! A router can associate with each datagram a given resource allocation 



Abstraction called a FLOW

A FLOW is a path through an internet along which intermediate routers guarantee a certain level of quality of service

© MMII JW Ryder

CS 428 Computer Networking

21

Traffic Class 

 

FLOW LABEL in the base header contains a label that routers use to map a datagram to a certain specific flow and priority Flows can also be used within an organization to manage network resources Example 

Two applications that need to send and receive video can establish a flow over which the bandwidth and delay are guaranteed

© MMII JW Ryder

CS 428 Computer Networking

22

IPv6 Extension Headers Base Header NEXT=TCP

TCP Segment

One Base Header Base Header NEXT=ROUTE

Route Header NEXT=TCP

TCP Segment

Route Header NEXT=AUTH

Auth Header NEXT=TCP

Two Base Headers Base Header NEXT=ROUTE

TCP Segment

Three Base Headers © MMII JW Ryder

CS 428 Computer Networking

23

IPv6 Fragmentation 







As with IPv4, IPv6 arranges for destination to perform re-assembly In IPv6 however, changes were made that avoid fragmentation by routers IPv4 requires intermediate routers to fragment any datagram that is too large for the maximum transfer/transmission unit (MTU) of network over which it must travel IPv6 fragmentation is end-to-end

© MMII JW Ryder

CS 428 Computer Networking

24

IPv6 Fragmentation 



No fragmentation done on intermediate routers Source which is responsible for fragmentation has two choices  

Use guaranteed minimum MTU (1280 octets) Perform Path MTU Discovery 

© MMII JW Ryder

Identifies minimum MTU along path to the destination CS 428 Computer Networking

25

IPv6 Fragmentation  

Either case, the source fragments data IPv6 fragmentation inserts a small extension header after the base header in each fragment

0

8 NEXT HEADER

16 RESERVED

24 FRAG. OFFSET

29 31 RS M

DATAGRAM IDENTIFICATION RS is set t 0 and reserved. M marks last fragment. ID unique for re-assembly. Fragments must be a multiple of 8 octets. © MMII JW Ryder

CS 428 Computer Networking

26