Physical Networks, Media and Interface Devices

Introduction

A computer network is physically realised through transmission media, connectors, interface electronics and interconnecting devices. Although modern networks are often discussed in terms of cloud services, software-defined networking and wireless mobility, every communication still depends on a physical path. Signals must be generated, transported, received, regenerated and interpreted. Poor selection or installation of physical components can cause attenuation, interference, unsafe working conditions and intermittent faults that are difficult to diagnose.

This chapter examines copper, optical-fibre and wireless media; their connectors and transceivers; Ethernet standards and duplex operation; and the functions and safe deployment of common interface and internetworking devices. The treatment is aimed at diploma students who need both conceptual understanding and practical installation skills.

Learning Objectives

After studying this chapter, you should be able to:

• compare copper, fibre-optic and wireless transmission media;
• identify common connectors, transceivers and interface devices;
• distinguish simplex, half-duplex and full-duplex operation;
• interpret major Ethernet naming conventions and standards;
• explain the roles of repeaters, hubs, bridges, switches, routers, gateways, modems and media converters;
• select devices and media for practical network scenarios;
• install and test physical network components safely;
• recognise limitations, security concerns and maintenance requirements.

1. The Physical Layer and the Communication Path

The physical layer carries binary information as electrical, optical or electromagnetic signals. It defines such properties as signal levels, modulation, timing, pin assignments, connectors, cable types and maximum transmission distances. The data-link layer, by contrast, organises bits into frames and controls access to a shared medium.

A typical wired path is:

Network interface controller → transceiver → connector → cable → patch panel → cable → switch port

A wireless path replaces the cable with antennas and a radio channel. In both cases, the quality of the path is affected by distance, interference, reflections, connector losses, environmental conditions and installation practices.

Important physical-layer terms include:

• Bandwidth: the frequency range or nominal data-carrying capability of a medium. In networking, it is commonly expressed in bit/s, although the engineering meaning of bandwidth is a frequency range measured in hertz.
• Bit rate: the number of transmitted bits per second.
• Attenuation: reduction in signal power as a signal travels through a medium.
• Interference: unwanted energy from other electrical, optical or radio sources.
• Crosstalk: undesired coupling between adjacent copper pairs.
• Latency: time taken for a signal or packet to travel and be processed.
• EMI: electromagnetic interference that can corrupt electrical signals.
• Noise margin: the difference between an acceptable signal level and the minimum level required for reliable interpretation.

For a point-to-point link, the total channel length is the sum of all permanent cable, patch-cord and connection sections. A link should not be judged only by its nominal cable length: connectors, bends, splices and environmental conditions also affect performance.

Figure: Principal characteristics of copper, fibre-optic and wireless media.

2. Transmission Media

2.1 Copper Media

The two principal network copper media are twisted-pair cable and coaxial cable.

Twisted pair contains insulated conductors twisted together to reduce electromagnetic coupling. Unshielded twisted pair (UTP) relies on balanced signalling and pair geometry. Shielded variants, such as FTP and S/FTP, add foil or braid protection and require correct bonding and grounding practices. A shield that is installed incorrectly can create rather than eliminate interference problems.

Common copper structured-cabling categories include Category 5e, Category 6 and Category 6A. Category 5e supports Gigabit Ethernet over appropriate channel installations. Category 6 provides improved high-frequency performance, while Category 6A is designed for 10-Gigabit Ethernet over the standard horizontal-cabling distance of up to 100 m when the complete channel meets the relevant requirements. The exact capability depends on cable construction, connectors, installation quality and the applicable standard.

Copper advantages include low cost, simple termination, widespread availability and the possibility of carrying power using Power over Ethernet (PoE). Limitations include attenuation, crosstalk, susceptibility to EMI and the need to observe bend radius, pulling tension and separation from power cables.

Coaxial cable has a central conductor, dielectric insulation, metallic shield and outer jacket. It offers good shielding and controlled impedance. It remains important in broadband access, cable television, radio systems and some legacy networks, although twisted pair and fibre dominate enterprise Ethernet horizontal and backbone cabling.

2.2 Fibre-Optic Media

Fibre carries information as pulses of light through a glass or plastic optical core surrounded by cladding. The cladding has a lower refractive index, allowing light to remain within the core through total internal reflection.

• Multimode fibre (MMF) has a relatively larger core and supports multiple propagation paths. It is commonly used within buildings and data centres at moderate distances.
• Single-mode fibre (SMF) has a much smaller core and supports long-distance, high-capacity transmission with low modal dispersion.

Fibre is resistant to electromagnetic interference, electrically nonconductive and suitable for building-to-building links, industrial environments and backbone networks. Its disadvantages include more expensive optical components, greater sensitivity to contamination and more specialised splicing and testing.

Fibre must not be sharply bent, crushed or pulled by its connector. Dust caps should remain fitted until connection. Never look into a fibre connector or active optical port; invisible infrared radiation may injure the eye. Inspecting and cleaning connectors with approved tools is essential because microscopic contamination can cause substantial loss.

2.3 Wireless Media

Wireless networks transmit through radio or, in some cases, infrared or microwave links. A radio channel is shared and affected by distance, obstacles, antenna orientation, reflection, absorption, competing transmitters and regulatory limits.

Important wireless concepts include:

• frequency: the number of oscillations per second, measured in hertz;
• channel: an allocated portion of spectrum;
• access point: a device that provides wireless station access to a wired or wireless network;
• SSID: the service-set identifier presented to users;
• roaming: movement between access points while maintaining service;
• received signal strength: the power level measured at the receiver, often in dBm;
• signal-to-noise ratio: the relationship between desired signal and background noise.

Higher frequency bands may provide more capacity but often have shorter effective range and poorer penetration through walls. Wireless deployment therefore requires a site survey, channel planning, appropriate transmit power, secure authentication and consideration of coverage overlap.

2.4 Media Comparison

3. Connectors, Interfaces and Transceivers

A connector provides a detachable mechanical and electrical or optical connection. An interface is the electrical, optical or logical boundary through which equipment communicates. A transceiver combines transmitter and receiver functions, converting data between an equipment interface and a physical medium.

Common interfaces include:

• RJ-45-style modular connectors: widely used for twisted-pair Ethernet. Strictly, many Ethernet plugs use an 8P8C modular connector, although “RJ-45” is common industry terminology.
• BNC: a bayonet connector used with some coaxial systems.
• F-type: common in broadband and cable television coaxial systems.
• LC: a small-form-factor fibre connector used extensively in switches and patch panels.
• SC: a larger push-pull fibre connector.
• ST: a bayonet-style fibre connector found in older installations.
• MPO/MTP: multi-fibre connectors used for high-density cabling.
• SFP and SFP+: pluggable transceiver form factors commonly supporting 1-Gb/s and 10-Gb/s applications, respectively, subject to the specific module and standard.
• QSFP families: higher-density pluggable modules used for technologies such as 40-Gb/s, 100-Gb/s and faster Ethernet implementations, depending on the variant.

A transceiver must match the switch or network interface, wavelength, fibre type, connector, reach and Ethernet standard. A module that physically fits may still be incompatible because of wavelength, coding, vendor restrictions or optical power budget.

A patch panel provides a structured termination point. It improves organisation and allows changes without repeatedly disturbing fixed horizontal cable. Patch cords should be of the correct category and length, and cable labels should identify both ends.

4. Duplex Modes and Ethernet Operation

Simplex communication travels in one direction only. Broadcast radio is an example. Half-duplex permits communication in both directions, but not simultaneously. Full-duplex permits simultaneous transmission and reception.

Early shared Ethernet used Carrier Sense Multiple Access with Collision Detection (CSMA/CD). A station listened before transmitting and detected collisions. With modern switched Ethernet, each link is normally point-to-point and full-duplex, so collisions and CSMA/CD are not normally relevant. A duplex mismatch—one end full-duplex and the other half-duplex—can cause late collisions, errors and severe performance degradation.

Autonegotiation allows compatible interfaces to advertise speed and duplex capabilities. Both ends should normally be configured for autonegotiation unless a documented compatibility requirement dictates otherwise. Manually forcing one end while leaving the other in an incompatible mode is a common troubleshooting error.

5. Ethernet Standards and Naming

Ethernet standards are developed primarily through IEEE 802.3. Names commonly indicate data rate and medium. For example, in 1000BASE-T, “1000” refers approximately to 1000 Mb/s, “BASE” indicates baseband transmission and “T” identifies twisted pair. In 100BASE-FX, “F” identifies fibre and “X” indicates the encoding family historically associated with the standard.

Examples include:

The standard’s maximum reach is not the same as a guarantee for every installation. Channel length, insertion loss, crosstalk, connector quality and transceiver specifications must all be considered. Standards also evolve, so design work should consult current IEEE, ISO/IEC, TIA and local regulatory documentation.

6. Network Devices and Their Roles

Repeater

A repeater operates at the physical layer. It regenerates or reshapes signals to extend a segment. It does not understand MAC addresses or IP packets. Repeaters can overcome signal degradation but do not reduce traffic and may extend a collision domain in legacy Ethernet.

Hub

A hub is essentially a multiport repeater. It repeats incoming bits out other ports, so all attached stations share the same collision domain and bandwidth. Hubs are largely obsolete because switches provide better performance, isolation and monitoring. If encountered in a legacy environment, they should be documented and replaced through a controlled plan.

Bridge

A bridge connects LAN segments and forwards frames using MAC addresses. It learns which addresses are reachable through which ports and filters traffic accordingly. Modern switches are multiport bridges implemented largely in hardware.

Switch

A switch is a multiport bridge. It learns source MAC addresses, stores them in a forwarding table and forwards frames to the appropriate destination port when known. Unknown unicast, broadcast and some multicast traffic may be flooded according to configuration.

Switches provide separate collision domains per port, but they do not automatically create separate IP subnets. VLANs can logically separate broadcast domains, while routing is required to communicate between IP subnets or VLANs. Managed switches may support VLANs, trunking, link aggregation, PoE, quality of service, port security, monitoring and access control.

Router

A router forwards packets between different IP networks. It examines network-layer addresses and selects a next hop using a routing table. Routers can perform static routing, dynamic routing, address translation, firewalling and traffic filtering, depending on the model. Unlike a basic switch, a router separates broadcast domains.

Gateway

A gateway is a general term for a device or software function that provides access to another network or translates between different protocols or architectures. A default gateway is the router to which a host sends traffic destined for another IP network. In a broader sense, application gateways may translate between unlike application protocols.

Modem

A modem—modulator-demodulator—converts digital data into a signal suitable for a carrier service and recovers data at the receiving end. Examples include DSL, cable and cellular modems. A modem may be integrated with a router, but its modem and routing functions remain conceptually distinct.

Media Converter

A media converter changes one physical medium or interface type to another, such as copper Ethernet to fibre Ethernet. It normally preserves Ethernet framing rather than routing IP traffic. Speed, duplex, wavelength, connector type and management features must be matched. Poorly selected converters can introduce duplex mismatches or fail to support required negotiation.

7. Selecting and Deploying Devices Safely

A professional installation follows a design and verification process:

1. Define requirements: identify data rate, distance, PoE demand, environmental conditions, redundancy, security and future expansion.
2. Select the medium: use copper for economical short access links, fibre for long or EMI-sensitive links, and wireless where mobility or difficult cabling justifies a shared medium.
3. Check compatibility: match standards, connectors, transceivers, wavelengths, cable categories and power requirements.
4. Plan topology: document racks, patch panels, cable routes, device locations, uplinks and power sources.
5. Install correctly: maintain bend radius, avoid excessive pulling force, separate data and power cables as required, and label both ends.
6. Configure devices: set management credentials, VLANs, addresses, duplex policy, firmware and security controls.
7. Test: use cable certification tools for copper, optical power-loss or OTDR testing for fibre where appropriate, and throughput, latency and error monitoring for active links.
8. Document and hand over: record serial numbers, port maps, test results, configurations and maintenance procedures.

Safety includes de-energising equipment when appropriate, using insulated tools, observing electrical and PoE limits, securing racks against tipping, maintaining ventilation, protecting fibre from shards, and following laser-safety procedures. Never connect unknown equipment to a production network without authorisation. Change default passwords, restrict management access, use secure protocols and securely dispose of replaced devices because configuration data may remain in storage.

8. Worked Application: Choosing a Campus Backbone

A training centre has two buildings 180 m apart. Copper Ethernet is already available inside each building, but the route between buildings passes near industrial motors. The centre requires a reliable 10-Gb/s inter-building link and expects future expansion.

Copper is unattractive because the distance exceeds common horizontal-cabling assumptions and the motor environment increases EMI risk. Wireless could avoid trenching, but it would introduce a shared radio channel, weather and alignment considerations, and potentially variable capacity. A fibre backbone is therefore the preferred solution.

A practical design would use a suitable single-mode or multimode fibre system based on the chosen transceiver reach, installed in appropriate conduit with slack and protection. Each building would receive a managed switch with compatible optical modules. Fibre patch panels would provide organised termination, and the existing copper access switches would connect to the backbone switches.

Before acceptance, technicians would inspect and clean connectors, verify polarity, test optical loss against the transceiver budget, confirm 10-Gb/s link negotiation and measure traffic performance. The design should include spare fibres, labelled pathways, documented port assignments and a replacement-module plan.

Outcome: the centre obtains a high-capacity, electrically isolated backbone while retaining economical copper for desk connections.

Learning points: distance and EMI can change the preferred medium; a physically compatible module is not necessarily technically compatible; testing and documentation are part of deployment, not optional afterthoughts.

9. Limitations, Governance and Contemporary Practice

No medium is universally best. Copper is economical but limited by distance and interference. Fibre offers excellent capacity but demands skilled handling and clean terminations. Wireless supports mobility but requires spectrum management and stronger attention to authentication, encryption and rogue-device detection.

Network operators should comply with electrical codes, building regulations, fire-stopping requirements, occupational health and safety rules, spectrum regulations and cabling standards applicable in their jurisdiction. Fibre pathways that cross fire compartments may require approved fire stopping. Outdoor installations may need lightning protection, grounding and surge protection.

Contemporary networks increasingly combine high-speed fibre backbones, copper PoE access, Wi-Fi access points and programmable managed switches. AI-assisted monitoring can help identify abnormal error rates, congestion or device failures, but it should support—not replace—engineering judgement. Automated systems must be governed to prevent inappropriate configuration changes, privacy violations and unverified fault responses.

Best-Practice Checklist

• Use standards-compliant cable, connectors and transceivers.
• Keep copper pairs twisted as close as practical to the termination.
• Maintain fibre cleanliness and bend-radius limits.
• Label, test and document every installed link.
• Prefer full-duplex switched links and investigate duplex mismatches.
• Separate management access from ordinary user access.
• Replace hubs and obsolete repeaters where feasible.
• Secure wireless networks with strong authentication and current encryption.
• Protect equipment physically, electrically and thermally.
• Retain test results and configuration backups for maintenance.

Chapter Summary

Physical networks depend on a chain of compatible media, connectors, transceivers and interconnecting devices. Copper twisted pair is economical and can deliver PoE over short structured-cabling links, but it is vulnerable to attenuation, crosstalk and EMI. Fibre provides long reach, high capacity and electrical isolation, but requires careful handling, cleaning and specialised testing. Wireless provides mobility and flexible deployment, while operating over a shared and variable radio channel that requires planning and security.

Connectors and transceivers must match the medium, speed, wavelength, reach and standard. Full-duplex switched Ethernet is the normal modern operating model; duplex mismatches cause serious errors. Repeaters regenerate signals, hubs repeat traffic to all ports, bridges and switches forward frames using MAC addresses, routers forward packets between IP networks, gateways translate or provide access between systems, modems adapt signals for carrier services, and media converters change physical interfaces without normally routing traffic.

Safe deployment requires requirement analysis, standards-based selection, correct installation, testing, security, regulatory compliance and complete documentation.

Key Terms

Attenuation: Reduction in signal power during transmission.

Bridge: A device that forwards frames between LAN segments using MAC addresses.

Crosstalk: Unwanted signal coupling between adjacent conductors.

Duplex: The direction and simultaneity of communication.

Ethernet: A widely used family of IEEE 802.3 LAN technologies.

Fibre-optic cable: Cable that carries information as optical signals.

Gateway: A device or function providing access to or translation between networks or protocols.

Hub: A multiport repeater that repeats signals to multiple ports.

Interface: A boundary through which systems exchange signals or data.

Media converter: A device that converts between physical network media or interfaces.

Modem: A device that modulates and demodulates signals for a communications service.

Multimode fibre: Fibre supporting multiple optical propagation paths.

Repeater: A physical-layer device that regenerates signals.

Router: A device that forwards packets between IP networks.

Single-mode fibre: Fibre designed primarily for one optical propagation mode and long-distance transmission.

Switch: A multiport bridge that forwards frames according to learned MAC addresses.

Transceiver: A combined transmitter and receiver.

Twisted pair: Copper conductors twisted to reduce interference and crosstalk.

VLAN: A logical separation of a switched network into distinct broadcast domains.

Wireless access point: A device that connects wireless stations to a network.

Review Questions

Short Answer Questions

1. What is the main difference between a repeater and a router?
2. State two advantages and two limitations of fibre-optic media.
3. Why can a duplex mismatch reduce Ethernet performance?
4. What is the function of a transceiver?
5. Why should fibre connectors be inspected and cleaned before installation?

Long Answer Questions

1. Compare copper twisted pair, fibre optic and wireless media for an enterprise network.
2. Explain how a managed Ethernet switch learns and forwards frames.
3. Describe the differences among a bridge, switch, router, gateway and media converter.
4. Explain the meaning of the components in the Ethernet designation 1000BASE-T.
5. Develop a safe and systematic procedure for installing and testing a new network link.

Analytical / Application Questions

1. A factory needs a 10-Gb/s link between two buildings near heavy electric motors. Recommend a medium and justify the choice, including testing and safety requirements.
2. A network has frequent errors after one switch port was manually set to full-duplex while its attached device remained in an automatic mode. Analyse the likely cause and propose corrective steps.
3. A small training laboratory needs mobility for students, wired connections for fixed desktop computers and a reliable backbone to a server room. Design a suitable combination of media and devices, explaining traffic separation, security and maintenance considerations.

Multiple Choice Questions with Answers

Question 1

Which device forwards frames primarily by using MAC addresses?

A. Router

B. Switch

C. Modem

D. Wireless antenna

Correct Answer: B. Switch

Question 2

Which medium is inherently immune to electromagnetic interference?

A. Unshielded twisted pair

B. Coaxial cable

C. Fibre-optic cable

D. Shielded twisted pair

Correct Answer: C. Fibre-optic cable

Question 3

What does full-duplex operation permit?

A. Transmission in one direction only

B. Transmission in both directions, but never simultaneously

C. Simultaneous transmission and reception

D. Transmission only after a collision

Correct Answer: C. Simultaneous transmission and reception

Question 4

In 1000BASE-T, the letter T identifies which medium?

A. Twisted pair

B. Terahertz radio

C. Single-mode fibre

D. Coaxial trunk

Correct Answer: A. Twisted pair

Question 5

Which device forwards packets between different IP networks?

A. Hub

B. Repeater

C. Router

D. Patch panel

Correct Answer: C. Router

Question 6

What is a primary reason for using a media converter?

A. To translate a domain name

B. To change one physical network medium to another

C. To assign IP addresses automatically

D. To encrypt application data

Correct Answer: B. To change one physical network medium to another

Question 7

Which fibre type is generally associated with long-distance, high-capacity transmission?

A. Single-mode fibre

B. Unshielded fibre pair

C. Multimode plastic cable only

D. Coaxial fibre

Correct Answer: A. Single-mode fibre

Question 8

What is a major limitation of a traditional hub?

A. It cannot repeat electrical signals

B. It creates a separate collision domain for every port

C. All connected stations share the collision domain

D. It can route between IP subnets

Correct Answer: C. All connected stations share the collision domain

Question 9

Which practice is safest when handling an optical connector?

A. Look directly into the connector to check for light

B. Clean it with any available cloth

C. Keep dust caps fitted until connection and use approved inspection tools

D. Bend the fibre sharply to reduce slack

Correct Answer: C. Keep dust caps fitted until connection and use approved inspection tools

Question 10

A link has one end configured for half-duplex and the other for full-duplex. What is the most likely result?

A. Improved throughput because both modes combine

B. No effect under all conditions

C. Errors and poor performance caused by a duplex mismatch

D. Automatic conversion to fibre

Correct Answer: C. Errors and poor performance caused by a duplex mismatch