CS402: Computer Communications and Networks

The Evolution of Computing Power

• The processing power in today's smartphones surpasses the ENIAC computer of 1946 several million times.
• Today's gaming consoles far exceed the computational capabilities of supercomputers from the 1980s.
• Common devices like smart thermostats and wearable technology have more computing power than early computer systems.
• Advanced digital cameras contain processors that outperform some mainframe computers from several decades ago.
• The growth rate in computing power has started to slow. However, advancements continue in areas like quantum computing and specialized hardware for tasks like machine learning, a specialized branch of artificial intelligence (AI).
• Moore's Law, which predicted that processor speed would double every 18 months, is under revision as we approach the physical limits of silicon-based semiconductors.

The Growth of Networks

• From Time Magazine, July 27, 1998:
• Number of years it took for radio to reach 50 million domestic listeners:
• 40
• Number of years it took for television and cable to reach 50 million domestic viewers:
• 13
• Number of years it took to the World Wide Web to get 50 million domestic users:
• 4

Brief History of Computers and Networks

• Before 1950: Early Computers
• Expensive, cumbersome, taking a huge amount of space, and operated only by experts
• Late 1950’s and 60’s: Batch Processing
• Smaller, more affordable computers
• Users would have to physically carry their work (e.g. punched cards), to the computer
• Single user only
• 1960’s: Interactive Processing Through Timesharing
• Terminals with no processing power connected directly to the computer
• Multiple users simultaneously
• 1970’s: Minicomputers and Microcomputers
• Own processing capacity, yet less expensive and more flexible than traditional mainframes
• 1980’s: Personal Computers, Distributed Systems, Networks.

Brief History of Computers and Networks

• 1960’s: Interactive Processing Through Timesharing
• Terminals with no processing power connected directly to the computer
• Multiple users simultaneously
• 1970’s: Minicomputers and Microcomputers
• Own processing capacity, yet less expensive and more flexible than traditional mainframes
• 1980’s: Personal Computers, Distributed Systems, Networks.
  

Chronology of the Internet Evolution

• 1964: Paul Baran wrote reports outlining packet networks
• 1969: First ARPANET nodes operational
• First two important applications: Telnet and FTP
• 1972: Distributed e-mail invented
• 1973: First non-US computer linked to ARPANET
• 1975: ARPANET transitioned to Defense Communications Agency
• 1980: TCP/IP experimentation begins
• 1981: New host was added every 20 days
• 1983: TCP/IP switchover complete
• 1986: NSFnet backbone created
• 1990: ARPANET Retired
• 1991: Gopher introduced
• 1991: WWW invented
• 1992: Mosaic Introduced
• 1995: Internet Backbone privatized
• 1998: The number of registered domain names exceeds 2 million
• 2000: The number of indexable web pages exceeds 1 billion

Uses of Computer Networks

• Business Applications
• Home Applications
• Mobile Users

Networks vs Distributed Systems

• Network:
• An interconnected collection of autonomous computers, where users are completely aware of the network's existence.
• Users must explicitly generate a change.
• Distributed System:
• An interconnected collection of autonomous computers, where all computers operate as a single, large virtual computer system.
• The existence of multiple computers is transparent to users.
• A distributed system is a software system built on top of a network. Example: WWW

Network Classification

• Local Area Networks
• Metropolitan Area Networks
• Wide Area Networks
• Wireless Networks
• Home Networks
• Internetworks

Network Classification

Classification of interconnected processors by scale

Type of Network Links (transmission technology)

• Broadcast links
• Many machines share a common link.
• Point-to-point links
• Two machines share a single connection link.

Switching Techniques: Circuit Switching

• A physical path between the sender and receiver is established. The path remains active until the “call” is completed.
• It can involve a long set-up time while the common carrier searches for the free path.
• No data is transmitted while the circuit is established.
• Example of Connection-Oriented service. Modeled after the telephone system.
• Advantages
• Low overhead (only small circuit ID needed)
• Packets arrive in sequence
• QoS is easier to implement
• Disadvantages:
• Long set-up times (high idle time)
• Non-resilient in case of router failure

Switching Techniques: Packet Switching

• The source node divides the data into packets and transmits the packets. Each packet has a header that includes the address of the destination node, as well as a sequence number, indicating the position of the packet in the message.
• The source node transmits to the first switching node in the network
• The switching node sends the packet to another node (or to the destination) based on the packet address. The process is repeated until the packet gets to the final destination.
• The destination node uses the sequence numbers to reassemble the packets in the correct order.
• Example of connectionless service, modeled after the Postal System. Also known as Datagram packet switching

Datagram Packet Switching

• Advantages:
• No setup times (minimum idleness)
• Flexibility in the face of congestion
• More robust in the event of router failures
• Disadvantages:
• High Overhead
• More processing power at each node (larger addresses, per packet routing)
• More processing at the end node (packets arrive out of order)
• Jitter (variation in packet delay)

Switching Techniques

• Virtual Call Switching:
• The first packet determines the route that all packets will follow.
• Combination of Circuit Switching and time-division multiplexing.
• Packets from different sources will be interleaved in the transmission.
• When you have multiple circuits in one wire, each circuit is called a virtual circuit.

The Concept of Layering: Divide and Conquer

• Divide the Network functions into logical layers.
• Each layer is composed of software and/or hardware modules that perform related network services.
• Each layer uses the services provided by the layer immediately underneath.
• Data to be transmitted must pass down through the layers of the source node to the communication medium (i.e. physical link).
• The data travels across the physical link and up through the layers of the destination node to the user. This is called end-to-end communications.
• More about Layering later in this course.

Network Architecture

• Each layer deals with messages (packets).
• Messages are generally limited to a maximum size.
• Each message contains a control or header information used to synchronize with the remote peer. The header contains “instructions” that tell the remote peer what to do with the message.
• Each message contains a data portion that contains arbitrary data.
• When layer N accepts data from layer N+1, it encapsulates the entire layer N+1 message in the data portion of the layer N packet.
• When the remote peer receives a message, it strips off the header information and passes only the data to the next higher layer.

Connection-Oriented and Connectionless Services

Six different types of service.

Networks vs Distributed Systems

• Network:
• An interconnected collection of autonomous computers, where users are completely aware of the network's existence.
• Users must explicitly generate a change.
• Distributed System:
• An interconnected collection of autonomous computers, where all computers operate as a single, large virtual computer system.
• The existence of multiple computers is transparent to users.
• A distributed system is a software system built on top of a network. Example: WWW

---

Source: Adopted from Eladio R. Cortes Ramos
This work is licensed under a Creative Commons Attribution-NonCommercial-ShareAlike 4.0 License.