Networking Foundations and Terminology

Introduction

Computer networking is the interconnection of computing devices so that they can exchange data, share resources, provide services, and support communication. A network may connect two computers in one room, thousands of devices across an organisation, or billions of systems across the Internet. Network devices and protocols provide the mechanisms that make this communication possible.

For a networking professional, understanding terminology is not merely a matter of memorising definitions. Terms such as bandwidth, latency, throughput, availability, topology, control plane, and management plane describe different aspects of network behaviour. Confusing them can lead to poor design decisions. For example, increasing a link’s bandwidth does not automatically eliminate delay, and a network with high nominal speed may still deliver low practical throughput because of congestion, errors, or inefficient protocols.

This chapter establishes the conceptual foundation required for later study of switches, routers, wireless access points, firewalls, addressing, routing, and network security.

Learning Objectives

By the end of this chapter, a learner should be able to:

1. Explain the purposes and benefits of computer networks.
2. Distinguish LANs, WANs, and related network scopes.
3. Describe common physical and logical topologies.
4. Differentiate bandwidth, latency, throughput, goodput, and availability.
5. Compare client-server and peer-to-peer architectures.
6. Explain the roles of standards bodies and RFCs.
7. Identify the principal roles of common network devices.
8. Distinguish the data, control, and management planes.
9. Apply foundational terminology to a practical network design problem.

1. Purposes of Computer Networks

Networks serve several related purposes:

• Resource sharing: Users can share printers, storage, applications, Internet connections, and specialised equipment.
• Communication: Email, instant messaging, voice, video conferencing, and collaborative platforms depend on networks.
• Centralised services: Authentication, name resolution, file storage, backup, and software distribution can be provided centrally.
• Distributed processing: Workloads can be divided among servers, clients, cloud services, and edge devices.
• Reliability and continuity: Redundant links and devices can allow services to continue when a component fails.
• Management and monitoring: Administrators can observe performance, apply policies, and troubleshoot systems remotely.
• Scalability: A well-designed network allows devices and services to be added without redesigning the entire environment.

A network is therefore both a technical system and an organisational resource. Its design must reflect requirements such as cost, performance, security, availability, geographical coverage, and regulatory obligations.

2. Network Scope and Classification

Network classifications often describe geographical or organisational scope. Boundaries are practical rather than absolute, and modern technologies can blur them.

Local Area Networks

A LAN normally connects devices under common administrative control within a limited area. Ethernet switches connect wired endpoints, while wireless access points connect wireless clients. LANs generally provide high capacity and relatively low delay, although performance depends on the medium, hardware, traffic load, and design.

Wide Area Networks

A WAN connects geographically separated sites. It often relies on telecommunications providers rather than cabling owned entirely by one organisation. WAN links may have greater delay, lower capacity, usage charges, or more complicated fault-management arrangements than LAN links. An organisation may connect offices using leased circuits, carrier Ethernet, software-defined WAN services, or encrypted tunnels across the Internet.

Internetworks

An internetwork is a network of networks connected by routers or other Layer 3 devices. The Internet is the most significant example. The term “Internet” with a capital I refers to the global system, while “internet” can describe any interconnected collection of networks.

3. Network Topologies

A topology describes how devices and links are arranged. Physical topology concerns actual cabling and equipment placement. Logical topology concerns how data flows, which may differ from the physical arrangement.

Bus Topology

In a bus topology, devices share a common communication medium. Early Ethernet networks used coaxial cable in a bus arrangement. A transmission was visible to devices sharing the medium, and termination was required at the ends. Bus networks are simple conceptually but are difficult to expand and troubleshoot; a cable fault can affect many devices. They are now largely obsolete in modern enterprise LANs.

Star Topology

In a star topology, each endpoint has a separate link to a central device, normally a switch or wireless access point. Modern switched Ethernet commonly uses a star or extended-star arrangement. A failed endpoint cable usually affects only one device, but failure of the central switch can affect all connected devices.

Ring Topology

In a ring, each node connects to two neighbouring nodes. Rings can provide predictable access and, with dual paths, resilience. However, a break or failed node may disrupt communication unless the technology supports bypassing or an alternate path. Ring arrangements have been used in metropolitan and provider networks, although many contemporary systems use routed or switched mesh designs instead.

Mesh Topology

A full mesh provides a direct link between every pair of nodes. For n nodes, the number of undirected links is:

Here, L is the number of links and n is the number of nodes. This formula assumes one direct bidirectional relationship between each pair and no duplicate links. A full mesh offers redundancy but becomes expensive as n increases. A partial mesh connects only selected pairs and is common in WANs, where important sites receive multiple paths.

Hierarchical and Hybrid Topologies

Enterprise networks often use a hierarchical design with access, distribution, and core functions. A small network may combine these functions in one device, while a larger network separates them for scalability and fault isolation. Hybrid topologies combine stars, partial meshes, wireless cells, and point-to-point links.

Topology selection should consider capacity, resilience, cost, physical constraints, security boundaries, and ease of maintenance rather than appearance alone.

4. Performance and Service Quality

Bandwidth

Bandwidth is the maximum rate at which a communication link can carry data under specified conditions. It is commonly expressed in bits per second: bit/s, kbit/s, Mbit/s, or Gbit/s. A byte contains eight bits, but decimal and binary prefixes must not be confused. Network manufacturers commonly use decimal SI prefixes, where 1 Gbit/s = 1,000,000,000 bit/s.

Bandwidth is a capacity, not a guarantee of application performance. A 1 Gbit/s interface may deliver less than 1 Gbit/s of useful application data because of protocol overhead, contention, hardware limitations, or congestion.

Latency

Latency is the time required for data to travel through a network or for a request to receive a response. It may include processing, queuing, transmission, and propagation components:

Each term is a time, normally measured in seconds or milliseconds. Transmission time is the time needed to place all bits of a frame onto a link:

S is the frame size in bits and R is the link rate in bit/s. Propagation time is approximately distance divided by signal propagation speed. These equations assume a single link and exclude additional protocol exchanges and retransmissions.

Round-trip time (RTT) is the time for a message to travel to a destination and for a response or acknowledgement to return. RTT is important to interactive applications and transport protocols. A high-bandwidth satellite link, for example, may carry substantial data but still have noticeable interaction delay because of propagation distance.

Throughput and Goodput

Throughput is the actual rate at which data is transferred over a network during a period. It is normally less than or equal to the available bottleneck capacity. If a system transfers D bits during an interval Δt, average throughput is:

D must be measured in bits and Δt in seconds, producing bit/s. Goodput is the rate of useful application data, excluding headers, acknowledgements, retransmissions, and other protocol overhead. Throughput can therefore be high while goodput is relatively low.

A useful engineering relationship is the bandwidth-delay product:

R is the link rate in bit/s and RTT is in seconds, so BDP is in bits. It estimates how much data could be in transit on a path. It does not state the achieved throughput; packet loss, congestion windows, receiver capacity, and protocol behaviour also matter.

Availability

Availability describes the proportion of time a service is operational and accessible. A basic measure is:

MTBF is mean time between failures and MTTR is mean time to repair or restore, both measured in the same time unit. This expression assumes a long-run repairable system with stable average failure and restoration behaviour. If MTBF is 99 hours and MTTR is 1 hour, then:

Availability is not the same as performance or security. A network can be available but slow, or available but compromised. Redundancy, preventive maintenance, monitoring, spare equipment, tested backups, and effective incident procedures can improve availability.

Jitter, Loss, and Congestion

Jitter is variation in packet delay and is particularly important for voice and video. Packet loss occurs when packets are discarded or never arrive. Congestion occurs when offered traffic exceeds available resources, causing queues, delay, loss, or reduced throughput. Quality of service mechanisms may classify and prioritise traffic, but prioritisation cannot create additional capacity.

5. Network Architectures

Client-Server Architecture

In a client-server architecture, clients request services and servers provide them. A web browser is a client, while a web server supplies web content. A domain name system server resolves names, and an authentication server verifies identities.

Advantages include centralised administration, consistent security policy, easier backup, and efficient sharing of specialised resources. Disadvantages include server cost, dependence on central services, possible bottlenecks, and the need for skilled administration. Modern systems frequently distribute a service across multiple servers or cloud regions to improve scale and resilience.

A server is defined by the service it provides, not simply by its physical form. A powerful desktop, virtual machine, container, or cloud instance may act as a server.

Peer-to-Peer Architecture

In a peer-to-peer architecture, participating devices can act as both clients and providers. Resources are distributed rather than controlled by one central server. Small workgroups may use peer-to-peer file sharing, and some content-distribution systems use peer-to-peer principles.

Peer-to-peer systems can be inexpensive and resilient because there is no single service host. However, administration, access control, backup, consistency, and auditing may be more difficult. The appropriate choice depends on scale, security, availability, and governance requirements.

6. Standards Bodies and RFCs

Interoperability requires shared standards. Important organisations include:

• IEEE: Develops standards for many local and metropolitan networking technologies, including Ethernet and wireless LAN standards.
• IETF: Develops Internet protocols through an open, technical process. Its work includes IP, TCP, routing protocols, and application protocols.
• ISO: Publishes international standards and is associated with the seven-layer Open Systems Interconnection reference model.
• ITU-T: Produces telecommunications recommendations concerning transmission, signalling, and network services.
• ICANN and IANA-related functions: Coordinate global identifiers such as domain names, protocol registries, and IP address allocations through the Internet ecosystem.

An RFC (Request for Comments) is a numbered technical publication, primarily associated with the IETF process. RFCs may define protocols, document best current practices, explain architecture, or record informational and experimental work. The name does not mean that every RFC is informal or merely a suggestion. Its status must be checked: some are Standards Track, some are informational, experimental, historic, or best current practice. Professionals should consult the relevant RFC and implementation documentation rather than relying on an isolated summary.

Standards improve interoperability, but implementations may differ in optional features, defaults, version support, and interpretation. Conformance testing and careful configuration remain necessary.

7. Roles of Network Devices

Network Interface Card

A network interface controller or card provides a device with network connectivity. It operates at the physical and data-link levels, creates or receives frames, and commonly has a hardware address such as an Ethernet MAC address.

Repeater and Hub

A repeater regenerates signals. A traditional hub repeats incoming signals to multiple ports and does not make forwarding decisions based on destination addresses. Hubs create one shared collision domain and have largely been replaced by switches.

Bridge and Switch

A bridge connects LAN segments and forwards frames using data-link addresses. A switch is a multiport bridge. It learns source MAC addresses, records them in a forwarding table, and forwards a frame toward the appropriate port when the destination is known. Unknown unicast, broadcast, and some multicast traffic may be flooded according to the switch’s rules.

Switches reduce contention by providing separate point-to-point links, but they do not automatically eliminate congestion or security risks.

Router

A router connects different IP networks and forwards packets using a routing table. It separates broadcast domains, selects next hops, and may perform functions such as address translation, filtering, or quality-of-service classification. Routing decisions are based on network-layer information and configured or learned routes.

Wireless Access Point

An access point connects wireless stations to a wired or wireless distribution system. It manages radio communication, authentication association processes, and frame bridging. Wireless performance is influenced by signal strength, interference, channel use, client density, and building materials.

Modem, Firewall, and Gateway

A modem adapts signals for a particular access technology, although modern provider equipment may combine modem, router, switch, and wireless functions. A firewall enforces traffic policies between trust zones. A gateway is a broad term for a device or service that connects unlike systems or serves as an exit point; in common LAN usage, the default gateway is usually a router interface used to reach remote networks.

8. The Three Network Planes

The plane model separates functions conceptually. Implementations may combine them physically, but the distinction helps engineers understand operation and troubleshooting.

Data Plane

The data plane, also called the forwarding plane, handles user traffic. It examines packet or frame information and forwards, filters, queues, or modifies traffic according to installed rules. Data-plane operations must usually be efficient because they occur for many packets.

Control Plane

The control plane determines how forwarding should occur. It includes routing protocols, neighbour discovery, spanning-tree calculations, and the creation of forwarding information. The control plane responds to topology changes and failures. If it fails while existing forwarding entries remain, traffic may continue temporarily, but new paths may not be learned and recovery may be impaired.

Management Plane

The management plane is used by administrators and management systems to configure, monitor, and maintain devices. Examples include secure shell, network management protocols, logging, telemetry, configuration automation, and time synchronisation. Management access should be authenticated, authorised, encrypted where appropriate, logged, and restricted to trusted networks.

These planes are related but not identical. A routing protocol is a control-plane function; forwarding a user packet is a data-plane function; changing a route through an administrative interface is a management-plane action.

9. Applied Case: Designing a Small College Network

Background

This illustrative case concerns a college with an administration building, laboratories, classrooms, and a library. It has 600 users, Internet access, online learning services, voice communication, and centrally managed authentication.

Problem

The existing network uses a flat LAN with several unmanaged switches. Users report slow file access, wireless disconnections, and difficulty reaching services during peak periods. There is no documented topology, and a single switch failure can interrupt a whole building.

Analysis

The requirements include reliable local connectivity, separation of student and administrative traffic, support for voice and video, controlled wireless access, and manageable Internet connectivity. The current flat design creates a large broadcast domain and provides limited visibility and resilience. The problem is not simply insufficient bandwidth; poor segmentation, weak management, and single points of failure also contribute.

Approach and Solution

The college adopts a hierarchical design. Managed access switches connect computers, phones, printers, and access points. Redundant distribution switches aggregate building traffic, while a core layer connects buildings and the data centre. Routers or Layer 3 switches provide inter-network routing. VLANs separate administration, students, voice, guest wireless, and network management traffic. Firewalls control traffic between sensitive zones and the Internet.

The control plane uses dynamic routing between major network components, while the data plane forwards ordinary user packets. The management plane uses authenticated administrative access, central logging, monitoring, configuration backup, and time synchronisation. Wireless access points are centrally managed and placed after a coverage survey.

Capacity planning measures peak throughput, latency, packet loss, and utilisation rather than relying only on interface labels. Critical switches and links are duplicated, and failover is tested rather than assumed.

Outcome

The redesigned network reduces broadcast scope, improves fault isolation, provides clearer security boundaries, and gives administrators evidence for troubleshooting. Some costs increase because managed equipment, redundant links, monitoring, and maintenance are required. The design therefore balances technical improvement with budget and operational capability.

Learning Points

1. A network is designed around services and risks, not only device count.
2. Segmentation can improve security and performance, but it requires routing and policy control.
3. High bandwidth does not guarantee low latency or high availability.
4. Documentation and monitoring are essential management-plane functions.
5. Redundancy is valuable only when failover mechanisms are correctly configured and tested.

10. Challenges, Governance, and Best Practice

Networks process personal, financial, educational, and operational information. Administrators should apply least privilege, protect management interfaces, retain logs proportionately, and follow applicable privacy and cybersecurity regulations. Monitoring should have a legitimate purpose, defined retention, access controls, and transparent governance.

Common technical challenges include legacy equipment, incompatible standards, address exhaustion, wireless interference, vendor-specific features, supply-chain risk, and rapidly changing cloud services. Artificial intelligence can assist with anomaly detection, capacity forecasting, and event correlation, but automated recommendations require validation. Poor data, opaque models, false positives, or automated changes can create outages and security problems.

Good practice includes documenting physical and logical topology, maintaining an asset inventory, using secure management protocols, separating user and management traffic, applying configuration control, monitoring performance and availability, testing backups and failover, updating firmware, and reviewing standards and vendor advisories. Engineers should distinguish measured facts from assumptions and record units, time periods, and test conditions when reporting performance.

Chapter Summary

Computer networks connect devices and services to support communication, resource sharing, centralised services, distributed processing, reliability, and administration. LANs cover limited areas, whereas WANs connect geographically separated sites; the Internet is a global internetwork. Physical and logical topologies influence cost, resilience, scalability, and fault isolation.

Bandwidth is link capacity, latency is delay, throughput is the achieved transfer rate, goodput is useful application data rate, and availability is the proportion of time a service operates. These measures are related but not interchangeable. Client-server systems centralise services, while peer-to-peer systems distribute service roles. Standards bodies and RFCs support interoperability, although implementations must still be tested and governed.

Network devices have distinct roles: switches forward frames within LANs, routers connect networks, access points provide wireless connectivity, firewalls enforce policy, and management systems configure and observe infrastructure. The data plane forwards traffic, the control plane determines forwarding behaviour, and the management plane supports administration. Effective network design combines these concepts with security, documentation, monitoring, resilience, and operational discipline.

Key Terms

• Bandwidth: Maximum data-carrying capacity of a link under specified conditions.
• Latency: Time taken for data or a response to travel through a network.
• Throughput: Actual rate of data transfer over a period.
• Goodput: Useful application-data rate excluding protocol overhead and retransmissions.
• Availability: Proportion of time that a service is operational and accessible.
• LAN: Local area network covering a limited site or area.
• WAN: Wide area network connecting geographically separated networks.
• Topology: Arrangement of network devices and communication links.
• Switch: Multiport bridge that forwards frames using data-link addresses.
• Router: Device that forwards packets between different IP networks.
• Firewall: System that applies traffic-control policies between security zones.
• Client: Device or application that requests a network service.
• Server: Device or application that provides a network service.
• Protocol: Agreed set of rules for communication.
• RFC: Numbered technical publication associated mainly with the IETF.
• Data plane: Functions that forward or process user traffic.
• Control plane: Functions that calculate and maintain forwarding information.
• Management plane: Functions used to configure, monitor, and maintain devices.
• RTT: Round-trip time for a message and response.
• Jitter: Variation in packet delay.

Review Questions

Short Answer Questions

1. What are two purposes served by computer networks?
2. Distinguish a LAN from a WAN.
3. Define bandwidth and throughput.
4. What is the primary forwarding role of a router?
5. What is an RFC?

Long Answer Questions

1. Compare star, ring, bus, and mesh topologies, including their advantages and limitations.
2. Explain why a high-bandwidth link may still provide poor application performance.
3. Compare client-server and peer-to-peer architectures in terms of administration, resilience, and security.
4. Explain the roles of IEEE, IETF, ISO, and ITU-T in networking.
5. Describe the data, control, and management planes and give an example of each.

Analytical / Application Questions

1. A company has three offices connected by WAN links. Recommend a topology and device arrangement that balances resilience and cost. Justify the design.
2. A network reports 1 Gbit/s interface capacity but users experience slow file transfers and high delay. Identify measurements and possible causes that should be investigated.
3. Design a basic segmentation and management approach for a college network containing student, staff, guest, voice, and administration traffic. Explain how switches, routers, firewalls, and monitoring systems would cooperate.

Multiple Choice Questions with Answers

Question 1

Which device primarily forwards frames using MAC addresses?

A. Router

B. Switch

C. Firewall

D. Modem

Correct Answer: B. Switch

Question 2

Which statement best distinguishes bandwidth from throughput?

A. Bandwidth is delay, while throughput is packet loss

B. Bandwidth is maximum capacity, while throughput is the achieved transfer rate

C. Bandwidth applies only to wireless networks

D. Throughput is always greater than bandwidth

Correct Answer: B. Bandwidth is maximum capacity, while throughput is the achieved transfer rate

Question 3

Which topology normally connects each endpoint to a central switch?

A. Bus

B. Ring

C. Star

D. Full mesh

Correct Answer: C. Star

Question 4

Which plane calculates routes and maintains forwarding information?

A. Data plane

B. Control plane

C. Management plane

D. Physical plane

Correct Answer: B. Control plane

Question 5

A network transfers 600 megabits in 3 seconds. What is its average throughput?

A. 50 Mbit/s

B. 180 Mbit/s

C. 200 Mbit/s

D. 1,800 Mbit/s

Correct Answer: C. 200 Mbit/s

Question 6

Which architecture centralises services such as authentication and file storage?

A. Client-server

B. Bus-based

C. Peer-to-peer only

D. Ring-based

Correct Answer: A. Client-server

Question 7

What does RTT measure?

A. The maximum link capacity

B. The variation in packet delay

C. The time for a message to travel to a destination and for a response to return

D. The percentage of successful repairs

Correct Answer: C. The time for a message to travel to a destination and for a response to return

Question 8

Which organisation is most closely associated with developing Internet protocol standards through RFCs?

A. IETF

B. IEEE

C. ISO

D. ITU-T

Correct Answer: A. IETF

Question 9

If MTBF is 90 hours and MTTR is 10 hours, what is availability using A = MTBF ÷ (MTBF + MTTR)?

A. 9%

B. 10%

C. 90%

D. 100%

Correct Answer: C. 90%

Question 10

Which statement about a firewall is most accurate?

A. It regenerates physical signals

B. It forwards every frame without inspection

C. It applies traffic policies between security zones

D. It only provides wireless connectivity

Correct Answer: C. It applies traffic policies between security zones