Lecture#3

1

Lecture 03 – Overview of Internet Architecture

Instructor:Engr. Musfara Farooqui

University of Education, Township

2

Lecture Objective• Internet Service Providers and Internet

Backbones– ISP Categories– POPs and NAPs

• Delay and Loss in Packet Switched Networks– Types of Delay– Comparing Transmission and Propagation Delay– Queuing Delay and Packet Loss

• Protocol Layers and Service Models– Layered Architecture– The Internet Protocol stack

• History of Computer Networking and Internet

3

Internet Service Providers• What is an ISP?

– An ISP is an organization that connects business or residential customers to Internet (backbone).

– An Internet Service Provider (ISP) is a company that provides access to the Internet.

– Their customers can be businesses, individuals or organizations.

– The advent of ISPs has made connecting to the Internet an affordable and convenient option for general people

• Internet structure is roughly hierarchical• In the public Internet, access networks situated at

the edge of the Internet are connected to the rest of the Internet through a tiered hierarchy of Internet Service Providers (ISPs)

4

ISP Categories• ISP Categories

– Tier-1 ISPs (Internet Backbone)– Tier-2 ISPs– Tier-3 ISPs

• Backbone Providers / Tier-1 ISPs– These ISPs are nationwide or multinational organizations

that control Internet routing.– They often own significant pieces of backbone itself

• National Providers / Tier-2 ISPs– These ISPs buy capacity (bandwidth) and routing

services from backbone providers and run Points Of Presence (POP: location of access points to the Internet) across the country.

• Local Providers / Tier-3 ISPs– These ISPs operate in the same way as the national ISPs,

but on a smaller geographical area

5

Points of Presence (POPs)• POPs are private peering points of ISPs• Within an ISPs network, the physical location /

points at which the ISP connect to other ISPs are known as Points of Presence (POPs)

• A POP is simply a group of one or more routers in the ISP’s network at which routers in other ISPs can connect.

• The POP is in the ISP’s switch site or in a colocation space, the contents will always contain “access” equipment and an IP router.

• At the core of the POP is a router that acts as the central hub for routing within the POP and is also used to terminate high capacity connections.

6

Network Access Points (NAPs)

• NAPs are public peering points of ISPs• When two ISPs are directly connected to each

other, they are said to peer with each other.• The NAP can be owned and operated by either

some third-party telecommunications company or by an Internet backbone provider.

• NAPs exchange huge quantities of traffic among many ISPs

• Often a NAPs uses high speed ATM switching technology, with IP running on the top of ATM

7

Backbone Providers / Tier-1 ISPs• Tier-1 ISPs

– Also known as Internet Backbone– Exists at the center of the Internet Architecture – Directly connected to each of the other tier-1

ISPs– Connected to a large number of tier-2 ISPs and

other customer networks– International in coverage– Two tier-1 ISPs can also peer with each other by

connecting together a pair of POPs, one from each of the two ISPs.

– The trend is for the tier-1 ISPs to interconnect with each other directly at private peering points.

– Examples (e.g., UUNet, BBN/Genuity, Sprint, AT&T)

8

Internet structure: Tier-1 ISPs

Tier 1 ISP

Tier 1 ISP

Tier 1 ISP

Tier-1 providers interconnect (peer) privately

NAP

Tier-1 providers also interconnect at public network access points (NAPs)

9

National Providers / Tier-2 ISPs

• Tier-2 ISPs– Provides smaller coverage as compared to tier-1– National Coverage– Connect to one or more tier-1 ISPs– Connect to other tier-2 ISPs as well.– Tier-2 ISPs typically have regional or national

coverage and connects only to a few of tier-1 ISPs

– A tier-2 ISP is said to be a customer of the tier-1 ISP to which it is connected, and the tier-1 ISP is said to be a provider to its customer.

– The trend for tier-2 ISPs is to interconnect with other tier-2 ISPs and with tier-1 ISPs at NAPs

10

Tier 1 ISP

Tier 1 ISP

Tier 1 ISP

NAP

Tier-2 ISPTier-2 ISP

Tier-2 ISP Tier-2 ISP

Tier-2 ISP

Tier-2 ISP pays tier-1 ISP for connectivity to rest of Internet tier-2 ISP is customer oftier-1 provider

Tier-2 ISPs also peer privately with each other, interconnect at NAP

Internet structure: Tier-2 ISPs

11

Local Providers / Tier-3 ISPs

• Tier-3 ISPs– last hop (“access”) network (closest to end

systems)– Local Coverage– Below tier-2 ISPs are the lower-tier ISPs,

which connect to the larger Internet via one or more tier-2 ISPs

– Users and content providers are the customers of lower-tier ISPs and lower-tier ISPs are the customers of higher-tier ISPs

12

Tier 1 ISP

Tier 1 ISP

Tier 1 ISP

NAP

Tier-2 ISPTier-2 ISP

Tier-2 ISP Tier-2 ISP

Tier-2 ISP

localISPlocal

ISPlocalISP

localISP

localISP Tier 3

ISP

localISP

localISP

localISP

Local and tier- 3 ISPs are customers ofhigher tier ISPsconnecting them to rest of Internet

Internet structure: Tier-3 ISPs

13

Internet structure: network of networks

• a packet passes through many networks!

Tier 1 ISP

Tier 1 ISP

Tier 1 ISP

NAP

Tier-2 ISPTier-2 ISP

Tier-2 ISP Tier-2 ISP

Tier-2 ISP

localISP

localISP

localISP

localISP

localISP Tier 3

ISP

localISP

localISP

localISP

14

Delay Packet Switched Networks

• Considering what can happen to a packet as it travels from its source to its destination.– As a packet travels from one node to other

node (host or end system), it suffers from several types of delays at each node along the path

• Most important types of delays are:– Processing Delay– Queuing Delay– Transmission Delay– Propagation Delay

15

Types of Delay• Processing Delay

– The time required to process (examine the packet’s header and determine where to direct the packet) is part of the processing delay

– Processing delay in high-speed routers is typically on the order of microseconds or less.

– After this nodal processing, the router directs the packet to the queue that precedes the link to the next router.

– Processing Delay depends on the processing speed of a router.

16

Types of Delay• Queuing Delay

– At the queue, the packet experiences a queuing delay as it waits to be transmitted onto the link.

– The queuing delay of a packet will depend on the number of earlier-arriving packets that are queued and waiting for transmission across the link

– If queue is empty, and no other packet is being transmitted, the queuing delay will be zero

– If traffic is heavy and many other packets are waiting to be transmitted, the queuing delay will be long

– Thus, queuing delay depends on the intensity and nature of traffic arriving at the queue.

– Queuing delays can be in the order of microseconds to milliseconds in practice

17

Types of Delay• Transmission Delay

– It is the amount of time required to push an entire packet into the link

– The time taken by a transmitter to send out all the bits of a packet onto the medium

– Also called Store and Forward Delay– Node receives complete packet before

forwarding– Transmission Delay is directly proportional

to the length of the packet– Transmission delays are typically in the order

of microseconds to milliseconds in practice

18

Types of Delay• Transmission Delay

– Let us denote the length of the packet by L bits.

– Denote the transmission rate of the link from Router A to B by R bits/sec

– Transmission Delay (L/R) = Packet Length (L)

Transmission Rate (R)

– Example:• It takes 1 sec to transmit a 10,000 bits

packet onto a 10Kbps line. (10,000 / 10 x 1000 = 1)

R R R

L

A B

19

Types of Delay• Propagation Delay

– Time it takes a bit to propagate from one node to the next.

– The time required by a bit to propagate from the beginning of the link to the next router is called propagation delay

– The bit propagates at the propagation speed of the link which depends on the physical medium being used.

– It is typically in the range of:• 2 x 108 meters/sec to 3 x 108 meters/second

– In wide area networks, propagation delays are on the order of milliseconds

20

Types of Delay• Propagation Delay

– Propagation delay depends on the distance (d) between the two routers/nodes and the propagation speed (s) of the link.

Propagation Delay (d/s) = Distance b/w 2 Routers (d)

Propagation Speed (s)

21

Types of Delay• Total Nodal Delay (the delay at a single

router)– If we let dproc, dqueue, dtrans and dprop denote the

processing, queuing, transmission and propagation delays respectively, then the total nodal delay is given by:

dnodal = dproc + dqueue + dtrans + dprop

22

Queuing Delay• Queuing delay is most complicated and

interested delay as compared to other components of nodal delay (processing, transmission, propagation)

• Queuing delay can vary from packet to packet– Example: if ten packets arrive at an empty

queue, the first packet will suffer no queuing delay while the last packet will suffer large queuing delay

23

Queuing Delay• Queuing delay depends on:

– Average Rate at which the packets arrives at a queue (a = packets/sec)

– Transmission Rate of the link (R = bits/sec)– Nature of the incoming traffic (bursty/periodic)– Assume that all the packets are of equal length

say L bits– Then the average rate at which the bits arrive

at the queue will be La bits/sec

• Traffic Intensity = La/R– This ratio helps in estimating the extent of

queuing delay

24

Traffic Intensity• Traffic Intensity

– If La/R is > 1• It means that the average rate at which the

bits arrive at the queue exceeds the rate at which the bits can be transmitted from the queue.

• In this undesirable situation, the queue will tend to increase without bound and the queuing delay will reach to infinity!

– A golden rule in traffic engineering• “Design your systems so that the traffic

intensity is no greater than 1s”

25

Traffic Intensity• Traffic Intensity

– If La/R is > 1• If the traffic intensity is close to one, there

will be intervals of time when the arrival rate exceeds the transmission capacity and a queue will form

• As the traffic intensity approaches 1, the average queue length gets larger and larger

– If La/R is < 1• If the traffic intensity is close to zero, then

the packets arrivals are few and far between, and it is unlikely that an arriving packet will find another packet in the queue

• Average queuing delay will be close to zero

26

Traffic Intensity

Traffic Intensity (La/R)

Average Queuing Delay

0 1

27

Applets Resources• Computer Networking; A Top Down

Approach Featuring the Internet– Applet Resources

• http://wps.aw.com/aw_kurose_network_2/0,7240,227091-,00.html

– Queuing and Loss Applet• http://media.pearsoncmg.com/aw/

aw_kurose_network_2/applets/queuing/queuing.html

28

Packet Loss• In reality a queue has a finite capacity• As the traffic intensity approaches 1, a packet can

arrive to find a full queue.• With no place to store such a packet, a router will

drop that packet; that is the packet will be lost• The fraction of lost packets increases as the

traffic intensity increases• Thus, a node performance also includes the

probability of packet loss• A lost packet may be retransmitted on an end-to-

end basis, either the application or transport layer protocol.

29

End-to-End Delay• The total delay from source to destination

is referred to as end-to-end delay– Example:

• Suppose that the queuing delay is negligible as the network is uncongested, then the end-to-end delay between the source and destination having N-1 routers in between will be:

dend-end = N (dproc + dtrans + dprop )

R R R

L

30

Delays and Routes in the Internet

• Traceroute– A program that sends multiple special packets

towards the destination– As these packets work their way towards the

destination, they pass through a series of routers.

– When a router receives one of these special packets, it sends a short message back to the source.

– This message contains the name and address of the router

– http://www.traceroute.org– For Details: Consult Traceroute: RFC 1393– To Do: Explore the Netstat tracert

commands

31

Layered Architecture• Design Philosophy of Layered Architecture

– The complex task of communication is broken into simpler sub-tasks or modules

– Each layer performs a subset of the required communication functions

– Each layer relies on the next lower layer to perform more primitive functions

– Each layer provides services to the next higher layer

– Changes in one layer should not require changes in other layers

– Helps in troubleshooting and identifying the problem

32

Internet Protocol Stack

Application

Transport

Network

Data Link

Physical

33

TCP/IP Protocol Suite• Application Layer

– Responsible for supporting network applications– Protocols include: HTTP. SMTP, FTP etc.

• Transport layer (End-to-end Communication)– Two transport layer protocols (TCP and UDP)– Transports messages between client and server

applications• Network Layer (Host-to-host Communication)

– Routing of datagrams from one host to another– IP works on this layers

• Data link Layer (Node-to-node Communication)– Logical interface between end system and network– Examples: Ethernet, PPP, ATM and Frame Relay

technologies• Physical Layer

– Transmission medium– Signal rate and encoding

34

PDUs in TCP/IP

35

Some Protocols in TCP/IP Suite

36

History of Internet• In 1960s the telephone network was the worlds

most dominant communication network• Uses Circuit switching which is appropriate for

voice traffic by supporting constant data rates• With the increasing importance of computers, the

need for interconnecting different geographically dispersed computers was realized.

• Three research groups laid the foundations of packet switching notion for computers communications:– MIT (Leonard Kleinrock)– Rand Institute (Paul Baran)– National Physical Laboratory (NPL)

37

History of Internet• Idea of Packet Switching• Principles of Packet Switching were conceived in

1957 by Paul Baran and others.• 1961--- First Paper by him on Packet Switching• 1964--- First Book on Internet in which Idea of

Packet Switching was declared more efficient than Circuit Switching

• Paul Baran used first time Digital Computer Technology for Communication between Switching Networks and divided the data into “Message Blocks” and reassembled at destination with some error detection technique

• Dynamic Routing of these Message Blocks was also proposed by Baran

• 1968--- First Packet Switching Network was designed and Implemented

38

The Internet’s Infancy: 1960s

• DARPA (Defense Advanced Research Project Agency) was established as an outcome of the Sputnik1 launch in 1957 by NASA (National Aeronautics and Space Administration), formally known as ARPA

• Computers in the form of Network was visualized and Implemented for data communication by Taylor

• 1966--- First Wide Area Computer Network was developed

• 1967--- First Packet Switching Router in the form of IMP (Interface Message Processor) was proposed; about a size of refrigerator

• 1968--- BBN designed IMPs and established the protocols allowing IMPs to communicate with each other.

39

The Internet’s Infancy: 1960s

• 1969--- Network Working Group (NWG) was formed to ensure the stability of communication protocols. Steve Crocker wrote first minutes of meetings

• IMP1: The first node of the ARPANET – http://www.lk.cs.ucla.edu/LK/Inet/birth.html

• The IMPs (Interface Message Processors) connected both host computers and other IMPs and functioned to:– Receive data– Check for errors– Retransmit, if error exists– Route the packets– Verify that packet are sent to intended receivers

40

The Internet’s Infancy: 1960s

• This documents was called RFC (Request for Comments) to take suggestions from peoples; later it became a Standard

• NWG designed first host-to-host protocols for host to IMP and computer to computer communication

• 1969--- Device Drivers were proposed to enable communication between different operating systems and hardware

• The destination IMPs used hop-by-hop acknowledgements. Since the source systems were different, so a software had to be designed to enable them to communicate, which is called a device driver

41

The Internet Early Years: 1970s

• 1970--- NCP (Network Control Protocol) was designed; used Stop and Wait flow control.It was the first host-to-host communication protocol that is used between the ARPANET end systems

• 1972--- Idea of Open-Architecture Network was floated

• 1973--- TCP (Transmission Control Flow Control) was designed for data transmission and Checksum was used for error detection

42

The Internet Early Years: 1970s

• Protocol Stack

APPLICATION

NCP

DEVICE DRIVER

IMP

43

The Internet Growth Begins: 1970 - 1980s

• 1973--- Ethernet was proposed as a LAN Technology

• 1974 --- First Ethernet protocol was developed• 1978 --- IP was proposed for Addressing purposes• 1980--- TCP/IP Protocol Suite was designed• UNET: First TCP/IP product was introduced for

Ethernet• BSD (Berkeley Software Division) Unix Operating

System was introduced• 1st January 1983--- It was decided to replace

NCP to TCP/IP for all Networks that gives birth to INTERNET

• 1983--- FTP, SMTP, DNS were introduced

44

The Internet Growth Begins: 1980s

• UDP comes into play for Real time Applications like Voice and Video

• 1984--- USENET modified for Newsgroups• 1986--- All Super Computers were connected to

form a Backbone Network called NSFNET which started from 56Kbps and in 1988 was converted to T1 Line I.e., 1.544Mbps

• 1988--- First Internet Worm was invaded effecting around 60,000 Hosts

• 1992--- WWW was created by Berners-Lee who also created First Web Server and Browser (Also designed HTTP later)

• 1993--- Clinton received first email at [email protected]

• 1993---- First Real Web Browser called MOSAIC was introduced

45

Internet Privatization: 1990s• 1994--- E-business started at Internet

• NSFNET decided to Privatize Internet by creating 4 NAPs (Network Access Points) and giving permission to ISPs to connect to NAPs

• 1995--- NSF Created High Speed Backbone Network Service to provide high-bandwidth connectivity (155 to 622Mbps) among NSF’s SCCs (Super Computer Centers)

• Internet2 was Created by Connecting all Top 100 Universities to these SCCs via GigaPOPs (Gigabits point of presence)

• Internet2: It is Hybrid Network whose Members are Major Universities and Research Organizations.

• Several Access Speed Transitions from 56Kbps Modem to ISDN (64-128Kbps), DSL Asymmetric Service to Cable Modems etc.

46

References• Computer Networking; A Top Down Approach

Featuring the Internet– 3rd Edition: Chapter 1, Jim Kurose and Keith Ross

• Data and Computer Communications– 7th Edition, William Stallings

• Data Communications and Networking– 3rd Edition, Behrouz A. Forouzan

• Data Communications and Computer Networks– Curt M. White

• Computer Networks– 4th Edition, by Andrew S. Tanenbaum

• Note: Slides are adapted from the companion web sites of referenced books.