VB Lecture 1 - American University of Beirut
Download
Report
Transcript VB Lecture 1 - American University of Beirut
CVEV 118/698
Internet
Lecture 1
Prof. Mounir Mabsout
Elsa Sulukdjian
Walid El Asmar
History
1957, USSR launches the first artificial satellite
called Sputnik I.
1958, President Eisenhower creates the
Advanced Research Project Agency (ARPA)
inside the Pentagon.
ARPA’s role is to initiate and fund high-tech
research projects.
History (cont’d)
1962, Joseph Licklider and Wesley Clark publish "OnLine Man-Computer Communication“.
In this visionary paper, the authors described how the
spread of programs and information among a large
number of computers connected by a universal
network would create a system more powerful than
could be built by any organization.
Immediately interested by this idea, the ARPA chooses
Licklider to head the Information Processing
Techniques Office (IPTO), created in October 1962.
Technical Innovations
Two technical innovations were necessary to
enhance the development this network
technology:
– The Packet Switching concept, developed
independently by Paul Baran (RAND) and
Leonard Kleinrock (MIT).
– The Distributed Network architecture,
studied by Paul Baran.
Packet Switching
In early 60’s, data transmission methods used in
telecom were very inefficient.
Data was sent from a pole to another in one batch,
through analog circuits.
The line between the two poles had to remain open
throughout the time of transmission.
Any error would alter the integrity of the whole
message, and require to restart the transmission
process.
Packet Switching (cont’d)
The concept, proposed by Baran and Kleinrock,
consists of dividing each message into several
independent packets.
Each packet can travel along a different path of the
network, through intermediary nodes.
Upon arrival to the destination of all packets, the
message is finally reassembled.
Communication is made digital.
The Network
Paul Baran compared three network architectures:
– The centralized network, where all hosts are directly connected
to a central unit.
– The decentralized network, with several super-nodes, each taking
care of a few hosts.
– The distributed network, where there would be NO central
authority. All nodes, now of equal status, become potential relays
in the transmission of an information.
The Network (cont’d)
The distributed network itself is assumed to be unreliable
at all times.
Every node would be equal in status to all other nodes of
the network, with its own authority to originate, pass and
receive messages.
A packet sent would be tossed from node to node like a hot
potato, more or less in the direction of its destination.
This transmission process might first appear very hazardous
and inefficient.
However, even if big parts of the network were blown away,
the packets would still stay airborne, and wind their way to
their destination, through whatever nodes happened to
survive.
ARPAnet on the way !
In 1966, the packet switching had been successfully
tested, and the concept of a distributed network
adopted.
The IPTO, under the direction of Robert Taylor,
allocated major funds to launch the network project that
would lead to the creation of the Arpanet. Lawrence
Roberts is chosen as the head of the team.
At the very early conceptual stage, the project had been
oriented toward the creation of a military network that
could survive partial damage, after an enemy attack.
However, the military side was never predominant during
the development of the network. Although it was financed
by the Pentagon, the Arpanet was never “classified”.
ARPAnet on the way !
1969, Arpanet is finally functional, as four poles are
connected in a network: UC-Los Angeles, the Stanford
Research Institute, UC-Santa Barbara and The University
of Utah.
1970, the Network Control Protocol (NCP) developed
by the Network Working Group (NWG) is ready.
As the project expands, new protocols (TCP/IP
replacing NCP), and new applications (Telnet, FTP,
etc…) are developed.
1975, divides the network in two: a public part (Arpanet)
and a military part (Milnet).
Internet on the way !
In the 70’s, other networks than Arpanet develop.
1981, the National Science Foundation (NSF) initiates the idea
of using the TCP/IP protocol with the regular telephone
connection.
1985, the NSFnet is created, connecting first the whole American
scientific community, to a single backbone.
Progressively, many countries connect to NSFnet starting from
Canada and northern European countries.
NSF encourages the interconnection of all existing networks.
Private clients start financing their own connection to the network.
1990, the first internet provider is born: “The World”.
What is Internet ?
The word “Internet” might appear a bit vague, it refers to:
– Physical links and network routers.
– Shared Resources.
– Transmission Protocols.
Internet provides with several services:
– The Electronic Mail to send messages.
– The World Wide Web to access/share information.
– NewsGroups or Discussion Groups.
– The File Transfer.
– Telnet to connect to distant machines.
– The Internet Relay Chat (IRC)
Get Connected
Technically, all that is needed to connect to internet is a
computer, a communication device (I.e. modem) and a
communication software (TCP/IP protocol).
You can be part of a Local Area Network (LAN); or
connect via a Provider.
Communication on internet is based on a “clientserver” model.
The client program formulates a specific request,
carried out by the server upon reception.
Get Connected
IP Addresses
How to find a specific machine on the net to address a
certain request ?
Each machine connected to internet has a unique IP
address, just like a phone number. (IPv4)
The IP address is a number of 4 × 8 bits. I.e.:
IP = 11001010 00001111 10101010 00000001
Corresponding in decimal to 202.15.170.1
Theoretically there are 2564 possible IP addresses.
Practically the number of arrangements is much smaller,
due to conventional restrictions, technical purposes and
even past waste.
IP Addresses (Cont’d)
IP addresses are organized into four different classes, each
corresponding to a certain type of network:
– Class A, where the most significant bit (MSB) is 0, and the seven next
bits are fixed:
0xxxxxxx yyyyyyyy… (I.e. IP < 127…)
– Class B, where the two MSB are 10, and the fourteen next bits are
fixed:
10xxxxxx xxxxxxxx yyyyyyyy… (I.e. IP < 192…)
– Class C, where the three MSB are 110, and the twenty-one next bits
are fixed:
110xxxxx xxxxxxxx xxxxxxxx yyyyyyyy… (I.e. IP < 224…)
– Class D
IPv4 is nearly saturated. IPv6 is a comparable system that relies 32 bytes
addresses.
Protocols
A packet is not only modeled in function of the data to
be transmitted and the destination IP address.
Protocols are communication standards. They are
organized in a Layer Model, where each layer is
independent.
Before being sent, a packet will pass through the layer
model, where each protocol will append to it a certain
header.(Encapsulation process)
The headers will contain information on the packet’s
application, transport, command, etc…
Upon reception of a packet, reading/understanding the
information starts with decoding the headers.
Encapsulation
Application
Application Data
Transport
Internet
Network
Access
TCP/IP Model
TCP hdr Application Data
IP hdr
TCP hdr Application Data
Frame hdr IP hdr
TCP hdr Application Data
IP and TCP/IP
The IP (Internet Protocol) is a fundamental part in the
transmission process.
Yet, IP is only involved at the emission level, and ignores all
the rest of the transmission process. (reception, integrity and
coherence of information, etc…)
The TCP (Transmission Control Protocol) completes IP by:
–
–
–
–
Establishing a connection.
Sequencing the packets.
Controlling the flux.
Managing errors.
The notation TCP/IP really means “TCP over IP” in
reference to the protocol layer model.
The Domain Name System
For practical reasons, there exists a transcription system called
DNS (Domain Name System), that translates IP addresses
into text addresses.
Usually an address contains a Name, and a Domain that can
be generic (com, net, org,…), or specific to a country (I.e. lb).
A URL (Uniform Resource Locator) should always start with
the following structure:
Protocol://server
i.e.: http://www.aub.edu.lb
The protocol can be http, ftp, etc… in reference to a certain
service.
Who Controls Internet ?
Well, nobody really does.
Still some benevolent groups inside the ISOC try to manage
its general technical orientation.
The ISOC (Internet Society), entirely composed of
benevolent, is in charge of the administration of the net.
The IETF (Internet Engineering Task Force) regularly
produce papers simply exposing some critical observations on
the net, or proposing to the IAB and the users new standards
and technical innovations.
The IAB (Internet Architecture Board) meets to adopt the
proposed standards, and approve the allocation of new
addresses and resources.
Who Controls Internet ?
What’s Next ?
Internet Programming…