Networks, Internet and Cloud Computing

Introduction

Information technology is increasingly organized around connectivity. A computer, smartphone, sensor, server, or application rarely operates in isolation; it exchanges data with other systems through networks. Networks support communication, resource sharing, distributed applications, electronic commerce, education, entertainment, public services, and cloud-based information systems.

A network may be as small as two devices connected in a home or as extensive as the global Internet. Understanding networks requires more than memorizing device names. IT professionals must analyze how devices are connected, how data is addressed and routed, how communication is standardized, how wireless systems differ from wired systems, and how cloud services deliver computing resources on demand.

This chapter develops these foundations and connects them to practical decisions involving architecture, availability, security, performance, and governance.

Learning Objectives

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

1. Distinguish major network types and topologies.
2. Explain layered communication models and important protocols.
3. Describe MAC, IP, port, and domain-name addressing.
4. Explain routing, wireless communication, Internet and Web architecture.
5. Compare cloud service and deployment models.
6. Explain edge computing and evaluate connectivity and availability options.
7. Analyze network and cloud choices in an applied organizational scenario.
8. Identify security, privacy, ethical, and governance concerns.

1. Network Fundamentals

A computer network is a collection of interconnected devices that exchange data using agreed communication rules. A network normally includes end devices, such as laptops and servers; intermediary devices, such as switches, routers, and access points; transmission media; protocols; and services.

A useful distinction is between a network and an internetwork. A network connects devices within one administrative or physical environment. An internetwork connects multiple networks. The Internet is a global internetwork based primarily on the Internet Protocol (IP) suite.

Network performance is commonly discussed using several related measures:

• Bandwidth is the maximum data-carrying capacity of a link, usually measured in bits per second.
• Throughput is the actual rate achieved by an application or transmission.
• Latency is the time required for data to travel from source to destination. It is often measured in milliseconds.
• Jitter is variation in packet delay, important for voice and video.
• Packet loss is the percentage of packets that fail to reach their destination.
• Availability is the proportion of time a service is operational and reachable.

Bandwidth and latency are different. A high-bandwidth link can transfer a large amount of data per second while still having substantial delay. Conversely, a low-bandwidth link may respond quickly for small messages but perform poorly when transferring large files.

Network Types

Networks can be classified by geographical scope, ownership, or purpose.

The classifications overlap. A WLAN is a LAN using wireless transmission, while a VPN may operate across a WAN or the Internet.

Network Topologies

A topology describes how devices are arranged physically or logically.

• In a bus topology, devices share one main communication medium. It is simple but a cable failure or shared-medium contention can affect many devices.
• In a star topology, devices connect to a central switch or access point. It is common in modern LANs because individual link failures are relatively easy to isolate; failure of the central device, however, can affect the connected segment.
• In a ring topology, each device connects to two neighbours. Data may travel around the ring, although many modern systems implement logical rings over other physical designs.
• In a mesh topology, devices have multiple interconnections. Meshes offer redundancy but require more equipment, configuration, and management.
• In a hybrid topology, different topologies are combined, as in an enterprise network with star-based office LANs and partially meshed WAN links.

Topology selection involves a trade-off among cost, scalability, fault tolerance, manageability, and performance. A critical data-centre service may justify redundant switches and links even when a small office does not.

2. Layered Communication and Protocols

Networking is complex because many functions must cooperate. Layered models divide the problem into responsibilities with defined interfaces. The seven-layer OSI reference model contains physical, data-link, network, transport, session, presentation, and application layers. The practical Internet model is commonly represented using link, Internet, transport, and application layers, although textbooks and vendors may use slightly different layer groupings.

Layering supports interoperability and troubleshooting. For example, a failed cable is investigated differently from an incorrect IP route or a failed application authentication process. Data is encapsulated as it moves down a protocol stack: an application message becomes a transport segment or datagram, an IP packet, and then a link-layer frame. At the receiving host, the process is reversed.

Important Protocols

• Ethernet defines common wired LAN frame formats and operation.
• Wi-Fi, standardized by the IEEE 802.11 family, provides wireless LAN communication.
• IP provides logical addressing and packet forwarding across interconnected networks.
• TCP provides connection-oriented, reliable, ordered byte-stream delivery using acknowledgements, retransmission, flow control, and congestion control.
• UDP provides a lightweight datagram service without TCP's reliability mechanisms. Applications may use it when low overhead or application-managed recovery is useful.
• DNS translates domain names into resource records, commonly including IP addresses.
• DHCP automatically supplies configuration such as IP address, subnet mask, default gateway, and DNS server.
• HTTP and HTTPS support Web communication. HTTPS is HTTP protected by TLS, which provides encryption and server authentication when correctly configured.
• SMTP, IMAP, and POP3 support different aspects of email transfer and retrieval.
• SSH supports secure remote administration.

A protocol is not merely a software product. It is a specification governing message formats, sequence, meaning, error handling, and often security expectations.

3. Addressing and Routing

Several forms of address operate simultaneously.

MAC Addresses

A Media Access Control (MAC) address identifies a network interface at the data-link layer. Ethernet and Wi-Fi use MAC addresses for local delivery. A MAC address is not a universal end-to-end identity for an Internet service, and devices may use privacy or randomized MAC addresses, especially in wireless environments.

IP Addresses

An IP address identifies an interface within an IP network. IPv4 uses 32-bit addresses, commonly written in dotted decimal notation, such as 192.168.1.20. IPv6 uses 128-bit addresses, written in hexadecimal groups, such as 2001:db8::20. IPv6 was developed partly because the supply of globally routable IPv4 addresses is limited.

An address is interpreted together with a prefix length. In CIDR notation, 192.168.1.0/24 indicates that the first 24 bits identify the network prefix and the remaining 8 bits identify host positions. Under the traditional basic subnet interpretation, this provides 2⁸ = 256 address values, of which 254 are usable host addresses when the all-zero network address and all-one broadcast address are reserved. This convention does not apply identically to every special-purpose prefix or IPv6 design.

For a general IPv4 prefix with n host bits, the number of address values is:

Here, n is the number of host bits. The familiar usable-host calculation is:

The subtraction assumes a conventional IPv4 subnet in which one address identifies the network and one is the directed broadcast address. It should not be applied automatically to IPv6 or unusual point-to-point configurations.

Worked example. A /26 IPv4 subnet leaves 32 − 26 = 6 host bits. It contains 2⁶ = 64 address values and, under the conventional rule, 64 − 2 = 62 usable host addresses.

Private IPv4 ranges, such as 10.0.0.0/8 and 192.168.0.0/16, are intended for internal use and normally require network address translation (NAT) or another gateway function for access to the public Internet. NAT can conserve public addresses, but it can also complicate inbound connections, troubleshooting, and end-to-end visibility.

Ports and Names

A port number identifies a logical service endpoint within a host. An IP address identifies a host interface or network location; the combination of IP address and port helps identify an application endpoint. A domain name, such as example.edu, is a human-oriented identifier resolved through DNS. These forms of identification should not be confused: changing a device's IP address does not necessarily change its domain name, and a domain name may resolve to multiple addresses.

Routing

Routing determines how packets move between networks. A router examines a destination IP address and selects a next hop using its routing table. Routes may be manually configured or learned through routing protocols.

Interior Gateway Protocols operate within an administrative domain. OSPF and IS-IS are examples of link-state approaches; they build a view of network connectivity and calculate suitable paths. BGP is the principal inter-domain routing protocol of the Internet. It exchanges reachability information between autonomous systems and supports policy-based decisions, not simply the shortest physical path.

Routing is distinct from switching. A switch generally forwards frames within a local network using data-link information, whereas a router forwards packets between IP networks. Modern devices may combine both functions, but the conceptual distinction remains useful.

4. Wireless Communication

Wireless systems transmit information through radio waves rather than a dedicated physical conductor. Wi-Fi networks use access points and client stations. Cellular networks use geographically distributed radio cells connected to an operator's core network. Bluetooth is designed for short-range, relatively low-power personal connections. Satellite and microwave systems can provide connectivity where terrestrial cabling is difficult.

Wireless communication is affected by distance, obstacles, interference, channel utilization, antenna characteristics, and regulatory restrictions. The theoretical capacity of a noisy communication channel is often expressed by the Shannon capacity relationship:

C is the theoretical maximum information rate in bits per second, B is channel bandwidth in hertz, and S/N is the signal-to-noise power ratio expressed as a dimensionless ratio. The relationship assumes an idealized channel model and does not include all real-world protocol overheads, interference patterns, or implementation limitations. It demonstrates that capacity can be increased by bandwidth or signal quality, but not without practical limits.

Wireless security should use modern protections such as WPA2 or WPA3 where supported, strong unique credentials, updated firmware, network segmentation, and controlled administrative access. Open wireless networks expose users to eavesdropping and malicious access points. Public Wi-Fi should be treated as untrusted, with encrypted application protocols and, where appropriate, an organizational VPN.

5. The Internet and the World Wide Web

The Internet is the global system of interconnected networks that communicate using the TCP/IP family and related standards. It is not the same as the World Wide Web. The Web is one Internet application, based mainly on browsers, Web servers, URLs, HTTP or HTTPS, HTML, CSS, JavaScript, APIs, and supporting services.

When a user enters a URL, a typical sequence is as follows:

1. The client identifies the domain and requests DNS resolution.
2. DNS returns one or more addresses, possibly using caching and geographic or load-based selection.
3. The client establishes transport communication, often TCP or a modern protocol using UDP, depending on the service.
4. For HTTPS, TLS negotiates encryption and authenticates the server using certificates and trusted certificate authorities.
5. The browser sends an HTTP request.
6. A Web server, reverse proxy, application service, and possibly database produce a response.
7. The browser interprets the returned resources and renders the page.

A Web application is often divided into presentation, application, and data tiers. Content delivery networks cache static resources near users, reducing latency and origin-server load. Load balancers distribute requests among service instances. Caching improves performance but creates consistency and invalidation challenges.

The Web also creates security responsibilities. Input validation, authentication, authorization, secure session management, patching, logging, backups, and protection against common attacks are essential. Encryption in transit does not automatically protect data at rest or prevent an authorized user from misusing data.

6. Cloud Computing

Cloud computing is a model for providing shared configurable computing resources—such as networks, servers, storage, platforms, and applications—rapidly and often on demand, with measured usage. Its defining characteristics include self-service provisioning, broad network access, resource pooling, rapid elasticity, and measured service.

Cloud Service Models

The shared-responsibility model is central. A provider may secure the underlying data centre, while the customer remains responsible for identity management, data classification, application configuration, and appropriate access. The exact division varies by service.

Cloud Deployment Models

• Public cloud uses provider-operated resources shared among many customers through logical isolation.
• Private cloud is dedicated to one organization, although it may be operated internally or by a service provider.
• Hybrid cloud integrates private and public environments.
• Community cloud serves organizations with shared requirements, such as regulatory or mission needs.
• Multi-cloud uses services from multiple public-cloud providers. It may reduce dependence on one provider but increases architectural and operational complexity.

Cloud migration is not automatically cheaper. Costs may include storage, compute time, data transfer, managed services, licenses, support, monitoring, security tooling, and staff training. Elasticity can improve financial efficiency when demand varies, but poorly governed resources can continue running unnecessarily.

7. Availability, Resilience, and Connectivity

Availability concerns whether a service is usable when required. A common availability measure is:

Both uptime and downtime must use the same time unit. If a service is operational for 8,760 hours and unavailable for 8.76 hours during a year, availability is 8,760 ÷ 8,768 ≈ 99.90%.

Availability is not identical to reliability. Reliability concerns the probability that a system performs without failure for a specified period, while availability also considers how quickly failures are repaired or services restored. Redundancy, failover, backups, monitoring, disaster recovery, and tested procedures improve resilience.

Connectivity options include:

• Fibre or leased-line connections, which generally provide high capacity and stable service.
• Business broadband, which may be economical but offer less symmetrical capacity or weaker service guarantees.
• Fixed wireless and microwave, useful where cabling is difficult.
• Cellular connectivity, useful for mobility and backup links.
• Satellite connectivity, valuable in remote locations but often affected by latency, weather, or capacity constraints.

Organizations should assess service-level agreements, latency, packet loss, installation time, repair commitments, geographic diversity, security, and total cost. Two circuits that share the same physical duct or provider may not provide genuine redundancy.

8. Edge Computing

Edge computing places processing, storage, or analytics closer to data sources and users rather than sending every operation to a distant centralized cloud. Edge locations may include factories, retail branches, vehicles, telecommunications sites, or gateway devices.

Edge computing is valuable when low latency, local autonomy, bandwidth reduction, or data sovereignty is important. For example, a manufacturing system may detect a dangerous machine condition locally even if the wide-area connection fails. The cloud can still aggregate historical data, train models, and provide centralized management.

Edge systems introduce challenges: physical exposure, inconsistent connectivity, limited computing resources, distributed patching, synchronization, and expanded attack surfaces. A sound design defines which decisions must occur locally, which data must be retained, how devices authenticate, and how updates are verified.

9. Applied Case: A University Hybrid Network

This is an illustrative case. A university operates a main campus, two remote learning centres, residence halls, laboratories, and an online learning platform. Students use laptops and mobile devices, while laboratories contain sensors and specialized equipment. The university experiences congested wireless networks, occasional Internet outages, inconsistent identity management, and rising demand for online teaching.

Problem

The existing design uses a flat LAN in several buildings. Critical servers, student devices, laboratory equipment, and guest users share broad network segments. A single Internet connection serves the campus. Some laboratory data is sent continuously to a public cloud, causing congestion and delaying local safety alerts.

Analysis

The university needs stronger segmentation, resilient connectivity, improved wireless capacity, and a clear division between cloud and local processing. Critical services have different requirements: safety controls need low latency and local availability; learning services require scalable Internet-facing capacity; guest access requires isolation; administrative records require strict access control and compliance.

Approach and Solution

The university introduces a hierarchical campus design with redundant core switches, access switches, and diverse fibre paths. VLANs and routed security boundaries separate administration, laboratories, residence halls, voice services, and guests. Network access control authenticates managed devices and restricts unknown endpoints.

Two Internet providers are connected through resilient edge routers. Dynamic routing and health monitoring support failover, while essential services use tested backup procedures. Wi-Fi capacity is improved through a site survey, appropriately planned channels, modern authentication, and separate staff, student, laboratory, and guest networks.

Laboratory gateways perform immediate anomaly detection at the edge. Summaries and selected raw data are sent to a cloud platform for long-term analytics. The university adopts SaaS for collaboration, PaaS for selected student applications, and IaaS for systems requiring customized operating-system control. Central identity federation and multi-factor authentication are applied across services.

Outcome

The design reduces broadcast exposure, limits unauthorized lateral movement, preserves local laboratory operation during a WAN interruption, and distributes traffic across available links. Cloud elasticity supports periods of high online-learning demand. However, the university must continue monitoring costs, testing failover, patching edge devices, reviewing supplier contracts, and training users.

Learning Points

1. Network architecture should reflect business and safety requirements rather than convenience alone.
2. Segmentation reduces risk but does not replace endpoint security.
3. Cloud and edge computing are complementary, not mutually exclusive.
4. Availability claims are meaningful only when dependencies and recovery procedures are tested.
5. Identity, governance, and operational skills are as important as equipment.

10. Security, Ethics, and Governance

Network and cloud decisions affect confidentiality, integrity, availability, privacy, and accountability. Organizations should classify data, apply least privilege, use strong authentication, encrypt sensitive data in transit and at rest where appropriate, maintain logs, patch systems, test backups, and establish incident-response procedures.

Privacy requires attention to collection purpose, retention, access, cross-border processing, and lawful handling. Cloud contracts should address data location, breach notification, subcontractors, deletion, portability, audit rights, service levels, and termination assistance. Regulatory duties vary by jurisdiction and sector, so technical design must be reviewed with legal and compliance specialists.

Ethical concerns include digital exclusion, surveillance, biased automated decisions, opaque data practices, and unequal service quality. A cloud or network architecture that is technically efficient may still be unacceptable if it excludes users with limited connectivity or processes personal data without meaningful safeguards.

Contemporary Developments and Best Practices

Current networking trends include software-defined networking, zero-trust security, IPv6 adoption, Wi-Fi 6 and later generations, private cellular networks, intent-based automation, network telemetry, and confidential or privacy-enhancing cloud technologies. Artificial intelligence is increasingly used for anomaly detection, capacity forecasting, and operational assistance, but automated recommendations require human oversight, quality data, and protection against manipulated inputs.

Practical best practices include:

• Document logical and physical network diagrams.
• Use standardized configurations and controlled change management.
• Segment users, devices, servers, and sensitive workloads.
• Prefer strong identity-based access over shared credentials.
• Monitor latency, loss, utilization, authentication events, and service health.
• Test restoration and failover rather than assuming they work.
• Design for IPv6 awareness even when IPv4 remains necessary.
• Measure cloud consumption and assign ownership for resources.
• Select connectivity based on business impact, not advertised speed alone.
• Review suppliers, contracts, exit strategies, and data-handling practices.

Chapter Summary

Networks connect devices through media, intermediary systems, and protocols. LANs, WANs, WLANs, VPNs, SANs, and IoT networks differ in scope and purpose, while star, mesh, ring, bus, and hybrid topologies involve different cost and resilience trade-offs. Layered models make communication interoperable and support systematic troubleshooting.

MAC addresses support local delivery, IP addresses support logical internetworking, ports identify service endpoints, and DNS provides human-oriented naming. Routers select paths between networks, while switches generally forward traffic within local segments. Wireless systems provide mobility but are sensitive to interference, range, security, and shared capacity.

The Internet is a global internetwork, whereas the Web is one application built on it. Cloud computing provides elastic, measured access to shared resources through IaaS, PaaS, SaaS, and related models. Public, private, hybrid, community, and multi-cloud deployments involve different governance and operational consequences. Edge computing moves selected processing closer to data sources, often complementing rather than replacing cloud services.

Effective architecture balances performance, cost, availability, security, privacy, regulatory duties, and operational capability. Resilient solutions use redundancy, segmentation, monitoring, tested recovery, and appropriate connectivity. Technical decisions must therefore be evaluated as organizational and ethical decisions as well as engineering choices.

Key Terms

Network: Interconnected devices and systems that exchange data.

Protocol: Agreed rules governing communication.

LAN: Local area network covering a limited site or area.

WAN: Wide area network spanning large geographical areas.

Topology: Physical or logical arrangement of network components.

Bandwidth: Maximum data-carrying capacity of a link.

Latency: Time required for data to travel between endpoints.

MAC address: Data-link identifier associated with a network interface.

IP address: Logical address used for communication across IP networks.

CIDR: Classless Inter-Domain Routing notation for network prefixes.

Port: Logical service endpoint within a host.

DNS: Distributed system that maps names to network resource records.

Routing: Selection of paths for packets between networks.

TCP: Reliable, connection-oriented transport protocol.

UDP: Lightweight datagram transport protocol.

Cloud computing: On-demand model for shared, configurable computing resources.

IaaS: Cloud model providing virtualized infrastructure resources.

SaaS: Cloud model providing ready-to-use software applications.

Edge computing: Processing performed near data sources or users.

Availability: Proportion of time a service is operational and reachable.

Review Questions

Short Answer Questions

1. What is the difference between bandwidth, throughput, and latency?
2. Why is a star topology commonly used in modern LANs?
3. Distinguish an IP address from a port number.
4. What role does DNS play in Web access?
5. Define edge computing and identify one situation in which it is useful.

Long Answer Questions

1. Explain the purpose of layered networking models and describe the functions of at least four important protocols.
2. Compare IPv4 and IPv6 addressing, including the purpose of prefixes and the limitations of conventional IPv4 subnet calculations.
3. Discuss the main cloud service and deployment models and explain how shared responsibility affects security.
4. Explain how a Web request travels from a browser to an application and back to the user.
5. Evaluate the technical and organizational factors that should influence a university's connectivity and availability strategy.

Analytical / Application Questions

1. An organization has 120 devices that must be placed in one conventional IPv4 subnet. What is the smallest suitable prefix, and what assumptions must be stated when calculating usable hosts?
2. A remote clinic has unreliable broadband, needs immediate local operation for medical devices, and must upload selected records to a central platform. Propose a network, edge, cloud, and backup-connectivity design, justifying your choices.
3. A company proposes moving all systems to a public cloud to eliminate its data-centre costs. Critically evaluate the proposal using availability, security, compliance, connectivity, skills, vendor lock-in, and total-cost considerations.

Multiple Choice Questions with Answers

Question 1

Which device primarily forwards packets between different IP networks?

A. Ethernet hub

B. Router

C. Keyboard

D. Printer

Correct Answer: B. Router

Question 2

A conventional IPv4 /26 subnet contains how many usable host addresses under the standard network-and-broadcast assumption?

A. 32

B. 62

C. 64

D. 126

Correct Answer: B. 62

Question 3

Which protocol translates domain names into resource records such as IP addresses?

A. DNS

B. DHCP

C. SMTP

D. SSH

Correct Answer: A. DNS

Question 4

Which cloud service model usually gives the customer control over the guest operating system while the provider manages the physical infrastructure?

A. SaaS

B. PaaS

C. IaaS

D. Community cloud

Correct Answer: C. IaaS

Question 5

What is a principal advantage of edge computing?

A. It guarantees unlimited storage at no cost

B. It places selected processing closer to data sources

C. It removes the need for authentication

D. It replaces all wide-area networks

Correct Answer: B. It places selected processing closer to data sources

Question 6

Which statement best distinguishes the Internet from the World Wide Web?

A. The Internet is an internetwork; the Web is an application running over it

B. The Web is the physical cabling used by the Internet

C. The Internet is limited to Web browsers

D. They are two names for exactly the same system

Correct Answer: A. The Internet is an internetwork; the Web is an application running over it

Question 7

Which characteristic is most associated with TCP?

A. Reliable, ordered delivery

B. Only local MAC addressing

C. Domain-name resolution

D. Wireless encryption by itself

Correct Answer: A. Reliable, ordered delivery

Question 8

A service operates for 8,760 hours and is unavailable for 8.76 hours in a period. Its availability is approximately:

A. 90.0%

B. 95.0%

C. 99.0%

D. 99.9%

Correct Answer: D. 99.9%

Question 9

Which topology generally provides multiple alternative paths and high redundancy, at greater cost and complexity?

A. Bus

B. Mesh

C. Single-star

D. Point-to-point without redundancy

Correct Answer: B. Mesh

Question 10

Under the shared-responsibility model, which task commonly remains the customer's responsibility in a public cloud?

A. Maintaining the provider's building power system

B. Manufacturing the provider's servers

C. Managing customer identities and access permissions

D. Operating the provider's cooling equipment

Correct Answer: C. Managing customer identities and access permissions