Encoding, Multiplexing, Media and Switching

Introduction

A communication network transfers information by converting it into signals, carrying those signals through a transmission medium, and directing them toward the intended destination. The quality, capacity, cost and reliability of a network depend strongly on four design choices: how data is encoded, how several communications share a link, which medium carries the signal, and how intermediate devices switch traffic.

This chapter develops these foundations for practical networking. It distinguishes data from signals, explains digital and analog encoding, compares copper, fibre-optic and wireless media, and examines frequency-division, time-division and wavelength-division multiplexing. It then compares circuit, packet and message switching and applies the concepts to a realistic network-design problem.

Learning Objectives

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

1. Explain the relationship between data, signals, bandwidth, bit rate and baud rate.
2. Describe common digital-to-digital, digital-to-analog, analog-to-digital and analog-to-analog encoding methods.
3. Compare guided and wireless transmission media using performance, cost, security and installation criteria.
4. Explain multiplexing and distinguish FDM, TDM, WDM and statistical multiplexing.
5. Compare circuit switching, packet switching and message switching.
6. Select suitable encoding, media and switching approaches for practical network applications.

1. Data, Signals and Transmission Performance

Data is information represented in a form suitable for processing. It may be digital, such as a sequence of bits, or analog, such as speech represented by continuously varying air pressure. A signal is the electrical, optical or electromagnetic representation used to carry data.

A signal may be periodic or non-periodic. A periodic signal repeats a pattern and can be described by its amplitude, period, frequency and phase. Frequency is related to period by:

where f is frequency in hertz (Hz) and T is period in seconds. A sinusoidal signal can be represented as:

where A is amplitude, f is frequency, t is time and φ is phase in radians. This model describes an ideal sinusoid; real communication signals often contain many frequency components.

Bandwidth has two related meanings. In analog transmission it is the range of frequencies occupied by a signal, measured in hertz. For a signal extending from f₁ to f₂:

In digital networking, bandwidth is also commonly used informally to mean data rate, measured in bits per second. These meanings should not be confused.

The bit rate is the number of bits transmitted per second. The baud rate, or symbol rate, is the number of signal symbols transmitted per second. If each symbol represents n bits, then:

where R_b is bit rate in bits/s, R_s is symbol rate in baud, and M is the number of distinct signal levels or symbols. This assumes all symbols are equally usable and that coding overhead is ignored. For binary signalling, M = 2, so one symbol carries one bit.

Noise and Signal Quality

Noise is unwanted energy added to a signal. Common sources include thermal noise, electromagnetic interference, crosstalk and impulse noise. A useful quality measure is the signal-to-noise ratio (SNR):

where P_s is signal power and P_n is noise power. In decibels:

The power ratio must be dimensionless. A larger SNR generally permits more reliable communication.

For a noiseless channel with bandwidth B hertz and M discrete signal levels, the Nyquist maximum theoretical bit rate is:

For a noisy channel, the Shannon capacity is:

Here C is capacity in bits/s, B is bandwidth in hertz, and SNR is expressed as a linear power ratio, not decibels. These are theoretical limits under stated assumptions; practical systems operate below them because of imperfect components, coding overhead and interference.

2. Encoding Techniques

Encoding maps data into signal patterns. The choice affects synchronisation, bandwidth, error detection, electromagnetic emissions and receiver complexity.

2.1 Digital-to-Digital Encoding

Digital-to-digital encoding converts a bit stream into a digital electrical or optical signal.

Unipolar NRZ uses one voltage for a binary value and zero voltage for the other. It is simple but has a large direct-current component and poor clock recovery during long runs of identical bits.

Polar NRZ uses positive and negative voltage levels. In NRZ-L, the level represents the bit value. In NRZ-I, a transition represents one bit value while no transition represents the other. NRZ-I can handle some long runs better, but long sequences without transitions can still cause synchronisation problems.

Manchester encoding combines data and clock information: every bit contains a transition in the middle. The transition direction represents the data value. It provides strong self-synchronisation but requires approximately twice the signalling rate of basic NRZ for the same bit rate.

Differential Manchester always has a mid-bit transition for clocking; an additional transition at the beginning of the bit interval represents one of the bit values. It is robust when the polarity of the received signal may be reversed.

Bipolar AMI represents binary zeros by no pulse and alternating positive and negative pulses for binary ones. It reduces the DC component and can support limited violation-based error monitoring. Long sequences of zeros may require substitution schemes such as B8ZS or HDB3.

2.2 Block Coding and Scrambling

Line coding alone may produce long runs without transitions. Block coding adds controlled redundancy before line coding. In 4B/5B, each group of four data bits is mapped to a five-bit codeword selected to provide sufficient transitions. The efficiency is:

where k is the number of data bits and n is the transmitted coded bits. For 4B/5B, efficiency is 4/5 = 80%, excluding other protocol overhead.

Scrambling replaces undesirable patterns with carefully designed alternatives while preserving the average data rate. The receiver reverses the process. Scrambling is not encryption: it improves signal properties, not confidentiality.

2.3 Digital-to-Analog Modulation

Digital information may modulate an analog carrier when transmission occurs over a radio channel or a passband telephone channel.

• ASK varies carrier amplitude.
• FSK varies carrier frequency.
• PSK varies carrier phase.
• QAM varies both amplitude and phase and can carry multiple bits per symbol.

For M-ary modulation, each symbol can represent log₂(M) bits under ideal mapping. Higher-order modulation increases bit rate for a given symbol rate but requires better SNR because constellation points are closer together.

2.4 Analog-to-Digital Conversion

To carry voice or other analog information digitally, the signal is sampled and quantised. The sampling theorem states that a band-limited signal with highest frequency fₘₐₓ can be reconstructed if:

where f_s is the sampling frequency. This is a theoretical minimum; practical systems often sample above it and use anti-aliasing filters. Aliasing occurs when insufficient sampling makes different frequencies appear indistinguishable.

If each sample is represented using n bits and samples occur at f_s samples/s, the uncompressed bit rate is:

For example, 8,000 samples/s with 8 bits/sample produces 64,000 bits/s before framing or compression overhead. Quantisation introduces an approximation error because the continuous amplitude is mapped to discrete levels.

2.5 Analog-to-Analog Modulation

Analog information can modulate an analog carrier using amplitude modulation (AM), frequency modulation (FM) or phase modulation (PM). These techniques remain relevant in broadcasting and radio systems, although modern digital communications frequently use digital modulation and error-control coding.

3. Transmission Media

A transmission medium is the physical path or electromagnetic space through which signals travel. Media are divided into guided media, which confine signals to a physical path, and wireless media, which radiate signals through space.

3.1 Guided Media

Twisted-pair cable consists of insulated copper wires twisted to reduce electromagnetic interference and crosstalk. Unshielded twisted pair (UTP) is inexpensive and easy to install; shielded variants provide improved resistance to interference. Twisted pair is common in Ethernet access networks and telephone systems. Its limitations include attenuation over distance, susceptibility to electromagnetic conditions and lower capacity than fibre over long links.

Coaxial cable has a central conductor, dielectric insulation, metallic shield and outer jacket. Its shielding provides better noise resistance than ordinary UTP, and it supports relatively high-frequency signals. It is used in broadband access, CCTV and legacy Ethernet, although fibre and twisted pair have displaced it in many local networks.

Optical fibre carries pulses of light through a glass or plastic core surrounded by cladding. It offers very high capacity, low attenuation, electrical isolation and resistance to electromagnetic interference. Single-mode fibre uses a very small core and is suitable for long distances; multimode fibre supports multiple propagation paths and is generally used over shorter distances.

Fibre requires careful handling. Excessive bending can cause loss, connectors require cleanliness, and splicing and optical testing require specialised skills. Fibre is difficult to tap without creating detectable signal loss, but it is not automatically secure: optical interception and attacks on active equipment remain possible.

3.2 Wireless Media

Wireless systems use radio, microwave, millimetre-wave, infrared or satellite links. Radio waves can travel through or around obstacles, depending on frequency and environment. Microwave links often require line of sight and accurate antenna alignment. Infrared is typically short range and does not pass through opaque walls. Satellite systems provide wide geographic coverage but may introduce substantial propagation delay.

Wireless performance is affected by path loss, reflection, diffraction, scattering, interference, fading and the shared nature of the radio spectrum. Mobility is a major benefit, but wireless networks require careful authentication, encryption, channel planning and radio-frequency regulation.

3.3 Media Comparison

Media selection should consider total cost of ownership, not only nominal data rate. Labour, ducts, maintenance, energy, regulatory requirements, future expansion and environmental conditions may dominate the initial cable price.

Figure: Main categories and characteristics of guided and wireless media.

4. Multiplexing

Multiplexing combines multiple lower-rate signals onto one shared higher-capacity link. A multiplexer (MUX) combines channels; a demultiplexer (DEMUX) separates them at the receiving end. Multiplexing reduces infrastructure cost and allows a provider to share expensive transmission capacity.

Frequency-Division Multiplexing

In FDM, each channel occupies a separate frequency band simultaneously. Guard bands reduce interference between adjacent channels. FDM is suitable for continuous analog streams and radio systems. Its design must account for channel bandwidth, guard bands, filtering and non-linear distortion.

Time-Division Multiplexing

In synchronous TDM, each input receives a repeating time slot, even when it has no data. It is simple and predictable but can waste capacity. In statistical TDM, slots are assigned dynamically to active inputs; frames normally include addressing or channel-identification information. Statistical TDM improves utilisation but introduces variable delay and buffering requirements.

If N sources each produce R bits/s and a synchronous multiplexer adds h overhead bits per frame, with each frame carrying one unit from every source, the output rate is approximately:

where F is the number of frames per second. This formula assumes equal source rates and one contribution per source per frame.

Wavelength-Division Multiplexing

WDM applies the FDM idea to fibre: separate optical wavelengths carry independent channels through one fibre. Coarse WDM uses widely separated wavelengths; dense WDM supports many closely spaced channels. Optical amplifiers and wavelength-selective components allow long-haul capacity to scale without installing new fibre.

Other Multiplexing Concepts

Space-division multiplexing separates channels by physical paths or antennas. Modern cellular systems may use multiple antennas through MIMO techniques. Code-division multiple access assigns spreading codes to users so signals overlap in time and frequency but can be separated by correlation. Its operation depends on code properties, power control and receiver processing.

Multiplexing is not the same as switching. Multiplexing shares one link; switching selects a path through a network of links and nodes.

5. Switching Techniques

Circuit Switching

Circuit switching establishes a dedicated end-to-end path before communication. The path may reserve frequency bands, time slots or physical resources. The three conceptual phases are setup, data transfer and teardown.

Circuit switching provides predictable capacity and usually stable delay once established. It is appropriate for constant-rate traffic, but setup time and reserved capacity can be wasteful during silence or bursty data transfer. Traditional telephone networks are the classic example.

Packet Switching

Packet switching divides a message into packets. Each packet carries control information and is forwarded through network nodes. In a datagram network, packets are routed independently and may arrive out of order. In a virtual-circuit network, a logical path is established and packets normally follow it, although resources may not be permanently dedicated.

Packet switching efficiently supports bursty traffic because links are shared statistically. Its costs include header overhead, queuing delay, congestion, packet loss and possible reordering. Applications or transport protocols may need sequencing, retransmission and flow control.

A simplified store-and-forward delay for a packet crossing H links is:

where dₚ is propagation delay per link, dₜ is transmission delay, d_q is queuing delay and d_proc is processing delay. This expression assumes similar links and delays at each hop; real paths may differ. Transmission delay for a packet of L bits on a link of rate R bits/s is:

Propagation delay depends on distance and signal speed, not packet length:

where d is physical distance in metres and v is propagation speed in metres/s.

Message Switching

In message switching, the complete message is stored at an intermediate node before being forwarded. No dedicated path is required. This permits efficient link sharing but requires large buffers and can create substantial delay, especially for large messages. It is historically important and conceptually related to store-and-forward systems, although modern networks normally segment messages into packets.

Comparison

6. Worked Example: Choosing a Link and Estimating Delay

A college wants to connect two buildings 2 km apart. The link must support 1 Gb/s, tolerate electromagnetic interference from industrial equipment and remain useful for future expansion. Each packet is 12,000 bits, and the signal propagation speed in the cable is approximately 2 × 10⁸ m/s. Ignore queuing and processing delay.

Fibre is the preferred medium because it offers high capacity, electrical isolation and good long-distance performance. The propagation delay is:

The transmission delay on a 1 Gb/s link is:

For one direction across one link, the simplified delay is approximately 22 μs. A real design would add switching, framing, coding, serialisation and queuing delays. The example demonstrates why increasing bandwidth reduces transmission delay but does not eliminate propagation delay.

7. Applied Case: Network Upgrade for a Technical College

Background

This illustrative case concerns a technical college with three buildings, a central data centre, wireless classrooms and a small industrial laboratory. The existing network uses copper access cabling, a congested wireless system and a shared 100 Mb/s inter-building link.

Problem

Students report slow access to virtual laboratories and intermittent wireless service. The industrial laboratory introduces electromagnetic interference, while the administration requires better separation of student, staff and guest traffic. The budget permits staged investment rather than a complete replacement.

Analysis

The inter-building link is a bottleneck and is exposed to interference. Copper remains appropriate for short horizontal runs because it is inexpensive and easy to terminate. Wireless is essential for mobility but cannot be treated as a substitute for all wired capacity. Bursty application traffic favours packet switching and statistical sharing rather than permanently reserved circuits.

Approach and Solution

The college installs a fibre-optic backbone using redundant paths where practical. Single-mode fibre is selected for the longest outdoor route, while multimode fibre is considered for short equipment-room connections. Existing twisted-pair cabling is retained for classroom access after certification and replacement of defective segments.

Managed switches use packet forwarding and virtual LANs to separate students, staff, laboratory equipment and guests. Quality-of-service policies prioritise interactive teaching applications over bulk downloads. Wireless access points use centrally managed channels, strong authentication and modern encryption. The design uses WDM only if future backbone capacity justifies optical multiplexing equipment; otherwise, spare fibre strands provide a simpler expansion path.

Outcome

The backbone removes the principal capacity bottleneck, fibre reduces interference risk, and packet-switched Ethernet supports efficient sharing among unpredictable users. Network segmentation limits the consequences of compromised guest or student devices. Monitoring is added to measure utilisation, retransmissions, coverage and failures rather than relying only on user complaints.

Learning Points

1. A medium should be selected from environmental and lifecycle requirements, not speed alone.
2. Wireless improves mobility but introduces interference, security and capacity-planning challenges.
3. Multiplexing and switching solve different problems and may be used together.
4. Segmentation and monitoring are essential operational controls, not optional extras.

8. Professional Practice, Security and Contemporary Developments

Network engineers should document cable routes, connector types, test results, optical budgets, radio surveys and change history. Installation should comply with applicable electrical, building, spectrum and cabling regulations. Unauthorised spectrum use can interfere with safety-critical services and may violate national law.

Security must be considered at every layer. Fibre does not guarantee confidentiality; wireless signals extend beyond walls; copper may leak electromagnetic energy; and switching equipment can be misconfigured. Use authenticated access, encryption, secure management protocols, segmentation, firmware updates and physical protection of distribution rooms. Monitoring should respect privacy and retain only data justified by operational or legal requirements.

Contemporary networks increasingly combine high-capacity fibre, software-controlled packet switching, Wi-Fi, cellular systems, edge computing and cloud services. Higher-order modulation, adaptive coding, massive MIMO, coherent optical transmission and AI-assisted monitoring can improve capacity or operational efficiency. However, AI-based diagnosis should support—not replace—engineering validation: false alarms, biased training data and opaque decisions can cause incorrect changes to critical infrastructure.

Chapter Summary

Encoding converts information into signal patterns suitable for transmission. Digital line codes differ in transition density, bandwidth, DC balance and synchronisation performance. Modulation allows digital data to use analog carriers, while sampling and quantisation convert analog information into digital form. Sampling must satisfy the Nyquist condition for a band-limited signal, and practical systems require filtering and coding overhead.

Guided media include twisted pair, coaxial cable and optical fibre. Wireless media provide mobility and flexible deployment but face interference, fading, shared capacity and security challenges. Multiplexing combines several channels on one link using frequency, time, wavelength, code or space. Circuit switching reserves a path, packet switching shares network resources using packets, and message switching stores complete messages before forwarding them. Good network design evaluates capacity, delay, reliability, security, regulation, cost and future growth together.

Key Terms

Data: Information represented for processing or communication.

Signal: A physical representation used to carry data.

Bandwidth: Frequency range of a signal; informally, also link capacity.

Bit rate: Number of bits transmitted per second.

Baud rate: Number of signal symbols transmitted per second.

Noise: Unwanted energy that disturbs a signal.

SNR: Ratio of signal power to noise power.

Encoding: Mapping data into signal patterns.

Modulation: Varying a carrier to represent information.

Sampling: Measuring an analog signal at discrete time intervals.

Quantisation: Mapping sample amplitudes to discrete levels.

Guided medium: A physical path that confines a signal.

Wireless medium: A medium in which signals propagate through space.

Multiplexing: Combining multiple channels on one shared link.

FDM: Multiplexing by separate frequency bands.

TDM: Multiplexing by separate time slots.

WDM: Optical multiplexing using separate wavelengths.

Circuit switching: Communication using a reserved end-to-end path.

Packet switching: Forwarding data in independently handled packets.

Message switching: Store-and-forward transmission of complete messages.

Review Questions

Short Answer Questions

1. Distinguish bit rate from baud rate.
2. Why does Manchester encoding assist synchronisation?
3. State two advantages of optical fibre over copper cable.
4. What is the main difference between synchronous and statistical TDM?
5. Why can packet switching produce variable delay?

Long Answer Questions

1. Explain digital-to-digital encoding methods and compare their bandwidth and synchronisation characteristics.
2. Compare twisted pair, coaxial cable, optical fibre and wireless media for a campus network.
3. Explain FDM, TDM and WDM, including their uses and limitations.
4. Compare circuit, packet and message switching in terms of setup, delay, efficiency and buffering.
5. Explain sampling, quantisation and the practical meaning of the sampling theorem.

Analytical / Application Questions

1. A network carries 20 sources, each producing 2 Mb/s. Explain whether synchronous TDM or statistical TDM is more appropriate when sources are active only intermittently. State the trade-offs.
2. A 10,000-bit packet crosses three 100 Mb/s links. Ignoring queuing and processing, calculate the transmission delay per link and discuss why propagation delay must be added separately.
3. Design a media and switching strategy for a campus containing an interference-prone laboratory, mobile classrooms and a data centre. Justify your choices using capacity, security, reliability and cost criteria.

Multiple Choice Questions with Answers

Question 1

Which quantity measures the number of signal symbols transmitted per second?

A. Bit rate

B. Baud rate

C. Bandwidth

D. Propagation delay

Correct Answer: B. Baud rate

Question 2

What is a principal advantage of Manchester encoding?

A. It eliminates all noise

B. It requires no transitions

C. It provides regular transitions for synchronisation

D. It doubles the distance of every cable

Correct Answer: C. It provides regular transitions for synchronisation

Question 3

A signal has a highest frequency of 4 kHz. What minimum sampling frequency is required theoretically?

A. 2 kHz

B. 4 kHz

C. 8 kHz

D. 16 kHz

Correct Answer: C. 8 kHz

Question 4

Which medium is least affected by electromagnetic interference?

A. Unshielded twisted pair

B. Coaxial cable

C. Optical fibre

D. Wireless radio

Correct Answer: C. Optical fibre

Question 5

Which multiplexing method assigns separate frequency bands to channels?

A. FDM

B. TDM

C. WDM

D. Statistical coding

Correct Answer: A. FDM

Question 6

What is a typical weakness of synchronous TDM?

A. It cannot use digital signals

B. It may waste slots assigned to inactive sources

C. It requires optical fibre

D. It always loses packet order

Correct Answer: B. It may waste slots assigned to inactive sources

Question 7

Which switching method normally reserves an end-to-end path before data transfer?

A. Packet switching

B. Message switching

C. Circuit switching

D. Statistical multiplexing

Correct Answer: C. Circuit switching

Question 8

A 12,000-bit packet is sent over a 1 Gb/s link. What is its transmission delay?

A. 0.12 μs

B. 12 μs

C. 120 μs

D. 12 ms

Correct Answer: B. 12 μs

Question 9

Which statement best describes message switching?

A. It reserves a continuous physical circuit

B. It forwards individual bits without storage

C. It stores and forwards the complete message

D. It uses only wireless media

Correct Answer: C. It stores and forwards the complete message

Question 10

If each symbol can take one of 16 distinct values, how many bits can one ideal symbol represent?

A. 2

B. 4

C. 8

D. 16

Correct Answer: B. 4