l9cn : network layer: internet protocol

7/31/2019 L9CN : Network Layer: Internet Protocol

1/17

Internet Protocol

Copyright The McGraw-Hill Companies, Inc. Permission required for reproduction or display.20.1

2020--11 INTERNETWORKINGINTERNETWORKING

connectingconnecting networksnetworks togethertogether toto makemake ananinternetworkinternetwork oror anan internetinternet..

Topics discussed in this section:Topics discussed in this section:

Internet as a Datagram Network

Internet as a Connectionless Network

20.2

. n s e ween wo os s

20.3

. e wor ayer n an n erne wor

20.4


2/17

. e wor ayer a e source, rou er, an es na on

20.5

. e wor ayer a e source, rou er, an es na on con nue

20.6

Note

w c ng a e ne wor ayer n eInternet uses the datagram approach to

pac e sw c ng.

20.7

Note

ommun ca on a e ne wor ayer nthe Internet is connectionless.

20.8


3/17

Two Key Network-Layer Functions

analo :

packets from routersinput to appropriate

routing: process oflannin tri from

router output

source to dest

forwardin : rocess ofroute taken by packetsfrom source to dest.

getting through singleinterchange

routing algorithms

20.9

Interplay between routing and forwarding

routing algorithm

local forwarding table

header

value

output

link0100 3010101111001

221

10111

value in arrivingpackets header

23

20.10

Router Architecture Overview

Two key router functions: run routing algorithms/protocol (RIP, OSPF, BGP)

forwardingdatagrams from incoming to outgoing link

20.11

Physical layer:bit-level reception

given datagram dest., lookup output port

using forwarding table in input portmemor

a a n ayer:e.g., Ethernet

goal: complete input port processing at line

speed queuing: if datagrams arrive faster than

forwarding rate into switch fabric

20.12


4/17

Three types of switching fabrics

20.13

Switching Via Memory

First generation routers:

traditional computers with switching under direct

control of CPUpacket copied to systems memory

per datagram)

Input OutputMemory

Port Port

System Bus

20.14

Switc ing Via a Bus

datagram from input port memoryto output port memory via a sharedbus

us con en on: sw c ng speelimited by bus bandwidth

32 Gb s bus Cisco 5600: sufficientspeed for access and enterprise

routers

20.15

Switchin Via An Interconnection Network

overcome bus bandwidth limitations

Ban an networks other interconnection nets initialldeveloped to connect processors in multiprocessor

advanced design: fragmenting datagram into fixed, .

Cisco 12000: switches 60 Gbps through theinterconnection network

20.16


5/17

Output Ports

Bufferingrequired when datagrams arrive from fabric

Scheduling disciplinechooses among queueddata rams for transmission

20.17

Output port queueing

line speed

queueing (delay) and loss due to output port bufferoverflow!

20.18

How much buffering?

buffering equal to typical RTT (say 250

e.g., C = 10 Gps link: 2.5 Gbit buffer

ecen recommen a on: w ows,buffering equal to RTT C.

N

20.19

Input Port Queuing

Fabric slower than input ports combined -> queueingmay occur at input queues

Head-of-the-Line (HOL) blocking: queued datagram at

front of queue prevents others in queue from moving

queueing delay and loss due to input buffer overflow!

20.20


6/17

2020--22 IPvIPv44

mechanismmechanism usedused byby thethe TCP/IPTCP/IP protocolsprotocols..


Fragmentation

Checksum

Options

20.21

The Internet Network layer

Host, router network layer functions:

Routing protocolsIP protocol

Transport layer: TCP, UDP

path selection

RIP, OSPF, BGP

datagram formatpacket handling conventions

Networklayer

orwar ngtable

ICMP protocol

error reporting

router signaling

Link layer

physical layer

20.22

. os on o v n pro oco su e

20.23

. v a agram orma

20.24


7/17

IP datagram format

ver length

32 bitspro oco vers on

number

header length(bytes)

total datagramlength (bytes)

head.len

type ofservice

16-bit identifier

headerchecksum

time tolive

max numberremaining hops

fragmentation/reassembly

type of dataflgs

fragmentoffset

upperlayer

32 bit source IP address(decremented at

each router)

u er la er rotocol32 bit destination IP address

data(variable length,

to deliver payload to Options (if any) E.g. timestamp,

record routetaken, specifyhow much overhead

or UDP segment)

to visit.

20 bytes of TCP

20 bytes of IP

= 40 bytes + app

layer overhead

20.25

. erv ce ype or eren a e serv ces

20.26

Note

e prece ence su e was par oversion 4, but never used.

20.27

Table 20.1 Types of service

20.28


8/17

Table 20.2 Default types of service

20.29

Note

e o a eng e e nes e o alength of the datagram including the

ea er.

20.30

. ncapsu a on o a sma a agram n an erne rame

20.31

. ro oco e an encapsu a e a a

20.32


9/17

Table 20.4 Protocol values

20.33

Example 20.1

An IPv4 packet has arrived with the first 8 bits as

shown:

01000010The receiver discards the packet. Why?

Solution

There is an error in this acket. The 4 leftmost bits

(0100) show the version, which is correct. The next 4

bits (0010) show an invalid header length (2 4 = 8).

The minimum number of bytes in the header must be20. The packet has been corrupted in transmission.

20.34

Example 20.3

In an IPv4 packet, the value of HLEN is 5, and the

value of the total len th field is 0x0028. How man

bytes of data are being carried by this packet?

Solution

The HLEN value is 5 which means the total number of

bytes in the header is 5 4, or 20 bytes (no options).The total length is 40 bytes, which means the packet is

carrying20bytes of data (4020).

20.35

. ax mum rans er un

20.36


10/17

. s or some ne wor s

20.37

IP Fragmentation & Reassembly

network links have MTU(max.transfer size) - largestpossible link-level frame.

different link types, differentMTUs

fragmentation:

in: one large datagram

arge a agram v e(fragmented) within net

one datagram becomesseveral data rams

reassembled only at finaldestination

IP header bits used to

reassembly

identify, order relatedfragments

20.38

. ags use n ragmen a on

20.39

IP Fragmentation and Reassembly

ID=

offsetfragflaglength=

One large datagram becomes

several smaller datagrams

xamp e

4000 byte datagram

MTU = 1500 bytes

ID=x

offset

=0

fragflag

=1

length=1500

ID

=xoffset=185

fragflag=1

length

=1500

data field

offset =

ID

=xoffset=370

fragflag=0

length

=1040

1480/8

20.40


11/17

. ragmen a on examp e

20.41

Figure 20.12 Detailed fragmentation example

20.42

Reassembly

When the fragments arrive what strategy should be

used to reassemble them? the ma be out of order

Solution

1. The fist fra ment has an offset field value of zero.2. Divide the length of the first fragment by 8. The

second fragment has an offset value equal to that result.

. .

The third fragment has an offset value equal to that result.

4. Continue the process. The last fragment has a more bit value

of 0.

20.43

. axonomy o op ons n v

20.44


12/17

2020--33 IPvIPv66

TheThe networknetwork layerlayer protocolprotocol inin thethe TCP/IPTCP/IP protocolprotocol

suitesuite isis currentlycurrently IPvIPv44.. AlthoughAlthough IPvIPv44 isis wellwell

designed,designed, datadata communicationcommunication hashas evolvedevolved sincesince thethe

..

deficienciesdeficiencies thatthat makemake itit unsuitableunsuitable forfor thethe fastfast--

rowinrowin InternetInternet..


AdvantagesPacket Format

20.45

Initial motivation: 32-bit address space

soon to be completely allocated.Additional motivation:

header format helps speedprocessing/forwarding

header changes to facilitate QoS

IPv6 data ram format:

fixed-length 40 byte header

20.46

IPv6 Header (Cont)

Priority: identify priority among datagrams in flow .

(concept offlow not well defined).

Next header:identify upper layer protocol for data

20.47

Advantages of IPv6

1. Better address space

2. Better header format

3. New options

4. Allowance of extension

5. Support for resource allocation

6. Support for more security20.48


13/17

. v a agram ea er an pay oa

20.49

Figure 20.16 Format of an IPv6 datagram

20.50

Other Changes from IPv4

processing time at each hop

, ,indicated by Next Header field

ICMPv6:new version of ICMP

additional message types, e.g. Packet TooBig

multicast group management functions

20.51

Table 20.6 Next header codes for IPv6

20.52


14/17

Table 20.7 Priorities for congestion-controlled traffic

20.53

. r or es or nonconges on-con ro e ra c

20.54

Table 20.9 Comparison between IPv4 and IPv6 packet headers

20.55

. x ens on ea er ypes

20.56


15/17

Table 20.10 Comparison between IPv4 options and IPv6 extension headers

20.57

2020--44 TRANSITION FROM IPvTRANSITION FROM IPv44 TO IPvTO IPv66

BecauseBecause of of thethe hugehuge numbernumber of of systemssystems onon thethe

n erne ,n erne , e e rans onrans on roro vv oo vv cannocanno

happenhappen suddenlysuddenly.. ItIt takestakes aa considerableconsiderable amountamount ofof

fromfrom IPvIPv44 toto IPvIPv66.. TheThe transitiontransition mustmust bebe smoothsmooth toto

preventprevent anyany problemsproblems betweenbetween IPvIPv44 andand IPvIPv66

systemssystems..

Dual Stack

Tunnelin

Header Translation

20.58

Transition From IPv4 To IPv6

simultaneous

How will the network operate with mixed IPv4

Tunneling:IPv6 carried as payload in IPv4

20.59

. ree rans on s ra eg es

20.60


16/17

. ua s ac

20.61

. unne ng s ra egy

20.62

Tunneling

A B E F

IPv6 IPv6 IPv6 IPv6

tunnelLogical view:

Physical view:

A B E FC D

IPv6 IPv6 IPv6 IPv6IPv4 IPv4

Flow: XSrc: ADest: F


Src:BDest: E

Src:BDest: E

data data


data


data

A-to-B: E-to-F:v IPv6- -

IPv6 insideIPv4

- -IPv6 inside

IPv4

20.63

. ea er rans a on s ra egy

20.64


17/17

Table 20.11 Header translation

20.65

l9cn : network layer: internet protocol

Documents