Unit 1: Network basics - Subjective Questions

Data communication is the exchange of data between two or more devices through a transmission medium such as a cable, fiber-optic link, or wireless channel.

Its five fundamental characteristics are:

• Delivery: Data must reach the correct destination device.
• Accuracy: Data must be delivered without errors or unacceptable changes.
• Timeliness: Data must arrive within the required time. This is especially important for real-time applications such as voice and video communication.
• Jitter: Jitter refers to the variation in packet arrival time. Low jitter is required for smooth audio and video transmission.
• Security: Data should be protected against unauthorized access, alteration, and destruction.

A communication system is considered effective only when it satisfies these characteristics according to the requirements of the application.

A data communication system consists of the following five basic components:

• Message: The information to be communicated, such as text, numbers, audio, video, or images.
• Sender: The device that generates and transmits the message. Examples include a computer, mobile phone, or sensor.
• Receiver: The device that accepts and interprets the transmitted message.
• Transmission medium: The physical or wireless path through which data travels. Examples include twisted-pair cable, coaxial cable, optical fiber, radio waves, and microwaves.
• Protocol: A set of rules that governs communication between devices. It specifies data formatting, addressing, transmission timing, error control, and security.

All five components must work together for successful communication. For example, when an email is sent, the sender is the user's device, the receiver is the destination mail server or device, the message is the email content, the medium is the network connection, and protocols such as SMTP and TCP control the exchange.

Computers represent all information internally as binary digits, or bits. Different types of information are represented as follows:

• Text: Characters are encoded using standards such as ASCII, extended ASCII, or Unicode. Unicode supports a very large number of writing systems and symbols.
• Numbers: Numbers may be represented as binary integers, signed numbers, or floating-point values. The representation depends on the range and precision required.
• Images: An image is represented as a collection of pixels. Each pixel is assigned a binary value representing its color or intensity. Image quality depends on resolution and color depth.
• Audio: Sound is converted into digital form by sampling its amplitude at regular intervals. If the sampling rate is samples per second and each sample uses bits, the approximate bit rate is bits per second for a single channel.
• Video: Video consists of a sequence of images, often accompanied by audio. It requires large storage capacity and high transmission bandwidth.

Standard encoding schemes are essential because the sender and receiver must interpret the same binary patterns consistently.

Transmission mode describes the direction of data flow between two communicating devices.

• Simplex mode: Data flows in only one direction. The receiver cannot send data back to the sender. A television broadcast is a common example.
• Half-duplex mode: Data can flow in both directions, but not at the same time. Each device takes a turn transmitting. Walkie-talkies use this mode.
• Full-duplex mode: Data flows in both directions simultaneously. Telephone conversations and modern switched Ethernet connections are examples.

Full-duplex communication generally provides better performance, but it may require separate channels or techniques for handling simultaneous transmission.

Network design involves making decisions that ensure efficient, reliable, secure, and scalable communication. Important issues include:

• Performance: The network should provide acceptable throughput, bandwidth, delay, and response time.
• Reliability: It should continue operating despite link failures, device failures, or congestion. Redundancy and alternate paths improve reliability.
• Scalability: The design must support future growth in users, devices, applications, and traffic.
• Security: Authentication, authorization, encryption, firewalls, and secure protocols should protect resources and data.
• Cost: Equipment, installation, maintenance, energy, and administration costs must be balanced against performance requirements.
• Topology: The arrangement of devices and links, such as star, bus, ring, mesh, or hybrid topology, affects cost and fault tolerance.
• Manageability: Administrators should be able to monitor, configure, troubleshoot, and upgrade the network.
• Quality of service: Applications such as voice and video may require guaranteed bandwidth, low delay, and low jitter.

A good design balances these factors rather than optimizing only one of them.

Networks can be classified according to the geographical area they cover:

• Personal Area Network (PAN): Covers a very small area around an individual, usually a few meters. Bluetooth connections between a phone and wireless earbuds are examples.
• Local Area Network (LAN): Covers a room, building, office, school, or campus. LANs generally provide high data rates and are privately owned.
• Metropolitan Area Network (MAN): Covers a city or large metropolitan region. It may connect multiple LANs within that region.
• Wide Area Network (WAN): Covers large geographical areas such as countries or continents. It uses leased lines, satellite links, cellular networks, or other long-distance technologies.
• Internet: The Internet is a global interconnection of independent networks using common protocols, especially the TCP/IP protocol suite.

As geographical coverage increases, communication may involve more intermediate devices, longer propagation delays, higher management complexity, and greater dependence on service providers.

The client-server model and peer-to-peer model differ in how resources and services are organized.

Client-server model:

• A dedicated server provides services or resources to clients.
• Management, security, backup, and access control are centralized.
• It is suitable for large organizations and services such as web hosting, email, and database management.
• A server failure may affect many clients unless backup servers are available.
• It usually requires higher installation and administration costs.

Peer-to-peer model:

• Each device can act as both a client and a server.
• Resources are shared directly among participating devices.
• It is simple and inexpensive for small networks.
• Administration and security are decentralized.
• Performance and availability depend on the participating devices.

Thus, client-server networks provide centralized control and better scalability, while peer-to-peer networks provide simplicity and direct resource sharing.

A network protocol is a set of rules that controls communication between devices. Protocols enable devices from different manufacturers and running different software to exchange information correctly.

The main elements of a protocol are:

• Syntax: Defines the format, structure, order, and encoding of data fields in a message.
• Semantics: Defines the meaning of each section of a message and the actions that should be taken after receiving it.
• Timing: Defines when data should be sent, how quickly it should be transmitted, and how long a device should wait for a response.

Protocols may also specify addressing, error detection, flow control, connection establishment, termination, and security procedures. Examples include HTTP for web communication, TCP for reliable transport, IP for internetworking, and Ethernet for local network access.

Layered architecture divides the complex process of network communication into a sequence of smaller, well-defined layers. Each layer performs a specific group of functions and provides services to the layer above it.

The main advantages are:

• Reduced complexity: Each layer can be designed and studied independently.
• Modularity: A change in one layer does not necessarily require changes in other layers.
• Interoperability: Devices and software from different vendors can communicate when they follow common layer specifications.
• Simplified troubleshooting: Network problems can be analyzed layer by layer.
• Standardization: Protocols and services can be developed according to recognized interfaces.
• Ease of development: Different teams can work on separate layers simultaneously.

During transmission, data moves downward through the sender's layers, crosses the physical medium, and moves upward through the receiver's layers. Each layer may add control information called a header; this process is known as encapsulation.

The OSI model contains seven layers, arranged from the highest to the lowest as follows:

1. Application layer: Provides network services directly to user applications, such as web access, email, and file transfer.
2. Presentation layer: Handles data translation, encryption, decryption, compression, and format conversion.
3. Session layer: Establishes, manages, and terminates communication sessions between applications.
4. Transport layer: Provides end-to-end delivery, segmentation, reassembly, flow control, and error control.
5. Network layer: Provides logical addressing and routing of packets between different networks.
6. Data link layer: Provides framing, physical addressing, error detection, and access control over a single link.
7. Physical layer: Transmits raw bits through a physical medium and defines electrical, optical, mechanical, and radio specifications.

The OSI model is mainly a conceptual and educational framework. It helps explain how network functions interact and how protocols can be organized.

Encapsulation is the process of adding control information to data as it moves from the application layer toward the physical layer at the sender.

• The application layer creates the original data.
• The transport layer adds a header, producing a segment or datagram.
• The network layer adds a header, producing a packet.
• The data link layer adds a header and usually a trailer, producing a frame.
• The physical layer converts the frame into a stream of bits for transmission.

At the receiver, the reverse process is called decapsulation. Each layer removes and examines its corresponding header or trailer before passing the remaining data to the next higher layer.

A simplified sequence is:

Encapsulation allows each layer to add the information needed to perform its own function without requiring higher layers to understand lower-layer details.

The OSI model and TCP/IP suite are both used to describe network communication, but they differ in structure and practical use.

The OSI model has application, presentation, session, transport, network, data link, and physical layers. The commonly used TCP/IP model has application, transport, internet, and network-access layers. TCP/IP is widely implemented, while OSI remains valuable for learning, design, and troubleshooting.

The TCP/IP protocol suite is commonly described using four layers:

• Application layer: Provides services to user applications. Examples include HTTP, HTTPS, FTP, SMTP, DNS, and SSH.
• Transport layer: Provides process-to-process delivery. TCP offers reliable, connection-oriented communication, while UDP offers faster, connectionless communication without guaranteed delivery.
• Internet layer: Provides logical addressing and routing across interconnected networks. IP, ICMP, and routing protocols such as OSPF operate in or support this layer.
• Network-access layer: Handles communication over a particular physical network, including framing, hardware addressing, and transmission of bits. Ethernet, Wi-Fi, and related link technologies operate here.

Some descriptions divide the network-access layer into separate data link and physical layers, resulting in a five-layer representation. The layers cooperate through encapsulation to deliver data from an application on one host to an application on another host.

Different layers use different types of addresses because each layer identifies a different entity:

• Physical address: Used at the data link layer to identify a device on a local network. A MAC address is a common example. It is used for frame delivery within the same network.
• Logical address: Used at the network or internet layer to identify a host across interconnected networks. An IP address is a logical address used by routers for packet delivery.
• Port address: Used at the transport layer to identify a particular process or application on a host. TCP and UDP port numbers are examples.
• Specific or application address: Used at the application layer to identify a particular service or resource, such as an email address, domain name, or URL.

For example, when a web page is requested, the domain name identifies the destination service, the IP address identifies the destination host, the port number identifies the web application, and the MAC address identifies the next-hop device on the local link.

The three addresses operate at different layers and serve different purposes:

• Physical address: Usually assigned to a network interface and used for local delivery over a particular network. It is typically associated with the data link layer and is not normally used for routing across multiple networks.
• Logical address: Assigned to a host or interface for communication across interconnected networks. IP addresses belong to the network layer and can change when a device moves to another network.
• Port address: Identifies a process or service running on a host. It belongs to the transport layer and allows multiple applications to use the network simultaneously.

These addresses work together to deliver data to the correct device and the correct application.

Data delivery uses several addresses in sequence:

1. The application address, such as a domain name or URL, identifies the desired service or resource.
2. The domain name may be translated into a destination IP address by the Domain Name System.
3. The transport layer assigns a source port and a destination port so that the correct applications can communicate.
4. The network layer places the source and destination IP addresses in a packet and routers use the destination IP address to select a path.
5. On each local link, the data link layer uses a source and destination MAC address for delivery to the next device, often the next-hop router.
6. At the destination host, the transport layer uses the destination port to deliver the data to the correct application.

The logical destination IP address generally remains associated with the end host, while the data link addresses may change at every hop. This layered use of addresses allows communication across multiple interconnected networks.

Network topology describes the physical or logical arrangement of devices and communication links.

• Star topology: Every device connects to a central hub or switch. It is easy to install, manage, and expand. Failure of a single link usually affects only one device, but failure of the central device can affect the entire network.
• Bus topology: All devices share a common backbone cable. It is inexpensive and simple for small networks, but a backbone failure can stop communication and traffic collisions may reduce performance.
• Ring topology: Each device connects to two neighboring devices, forming a ring. Data travels around the ring, but a single break can disrupt communication unless a dual-ring or bypass mechanism is used.
• Mesh topology: Devices have multiple interconnected links. It provides excellent redundancy and fault tolerance, but installation and maintenance are expensive.

Modern LANs commonly use a physical star topology with switches because it provides good performance, flexibility, and ease of management.

The following performance measures are important in network design:

• Bandwidth: The maximum capacity of a communication channel, usually measured in bits per second. In analog systems, it may also refer to a frequency range.
• Throughput: The actual rate at which useful data is successfully delivered. It is usually less than the theoretical bandwidth because of overhead, congestion, errors, and protocol limitations.
• Propagation delay: The time required for a signal to travel from the sender to the receiver through the medium.
• Transmission delay: The time required to place all bits of a packet onto the transmission medium.

If a packet has length bits and the link rate is bits per second, transmission delay is:

For example, a larger packet increases transmission delay, while a higher link rate decreases it. Total network delay may also include processing delay, queuing delay, and propagation delay.

Connection-oriented communication establishes a logical connection before data transfer begins. The communicating devices may negotiate parameters, maintain state information, acknowledge received data, and terminate the connection after communication. TCP is an example. It provides reliable, ordered delivery but introduces connection-establishment and management overhead.

Connectionless communication sends data units independently without establishing a connection in advance. Each packet may be routed separately, and the network does not necessarily maintain information about the complete communication. IP and UDP are examples of connectionless protocols.

The choice depends on whether reliability or low overhead and speed is more important.

The transport layer provides communication between processes running on end systems. Its functions may include:

• Segmentation and reassembly of application data.
• Process-to-process delivery using port numbers.
• Multiplexing and demultiplexing of application streams.
• Flow control between sender and receiver.
• Error control, acknowledgments, and retransmissions when supported.
• Connection establishment and termination for connection-oriented protocols.

The network layer provides host-to-host delivery across multiple interconnected networks. Its functions include:

• Logical addressing using IP addresses.
• Routing and forwarding packets.
• Internetworking between different link technologies.
• Fragmentation and reassembly in systems that support it.

The network layer selects a path to the destination host, whereas the transport layer delivers data to the correct process on that host. TCP operates primarily at the transport layer, while IP operates at the network or internet layer.