4. Wireless Communication Theory and Spectrum Engineering

Introduction and Context

Wireless communication is the physical and mathematical foundation of the Internet of Things (IoT). Sensors, actuators, mobile devices, industrial controllers, satellites, autonomous systems, and edge-computing platforms depend on the transfer of information through an electromagnetic medium that is shared, variable, and regulated. Unlike a wired channel, a wireless channel is not a stable point-to-point connection. Its properties change with distance, frequency, terrain, weather, mobility, obstacles, interference, antenna orientation, and network loading.

Wireless-system design therefore requires more than selecting a radio standard. It requires coordinated decisions about propagation, frequency spectrum, modulation, coding, multiplexing, antennas, transmit power, receiver sensitivity, interference, energy consumption, and regulatory constraints. In IoT, these decisions are especially difficult because devices may be inexpensive, battery-powered, physically small, intermittently connected, and deployed in dense or inaccessible environments.

This chapter develops an integrated framework for wireless communication theory and spectrum engineering. It connects physical-layer models with engineering decisions: how far a signal can travel, how much information can be transmitted reliably, how spectrum can be reused, and how energy and interference can be controlled. The discussion is relevant to technologies including Wi-Fi, Bluetooth Low Energy, IEEE 802.15.4, LoRaWAN, cellular IoT, private 5G, satellite IoT, and emerging cognitive-radio systems.

Learning Objectives

After studying this chapter, the reader should be able to:

1. Explain the main mechanisms of radio-wave propagation and their effects on IoT links.

2. Distinguish analog and digital modulation, coding, multiplexing, and multiple-access techniques.

3. Construct and interpret a wireless link budget.

4. Relate antenna characteristics, receiver sensitivity, noise, and interference to system performance.

5. Apply Shannon capacity and practical spectral-efficiency concepts to wireless design.

6. Evaluate spectrum-allocation and cognitive-radio approaches.

7. Analyze trade-offs among range, throughput, reliability, latency, and energy consumption.

8. Critically assess technical, regulatory, security, ethical, and environmental issues in wireless deployment.

The Wireless Channel as a Communication System

A communication link consists of a source, transmitter, propagation channel, receiver, and destination. The transmitter maps information into a waveform; the channel modifies that waveform; and the receiver estimates the transmitted information from a noisy and often distorted observation. A useful baseband representation is

r(t) = h(t) * s(t) + n(t) + i(t),

where s(t) is the transmitted signal, h(t) represents the channel response, n(t) is thermal noise, i(t) is interference, and * denotes convolution. In a static narrowband channel, h(t) may be approximated by a complex gain. In a mobile or wideband channel, it varies with time and frequency.

Wireless performance is commonly described using received power, signal-to-noise ratio (SNR), signal-to-interference-plus-noise ratio (SINR), bit-error rate (BER), packet-error rate (PER), outage probability, latency, and energy per delivered bit. These measures are related but not interchangeable. A link may have high data rate but poor reliability, or excellent reliability at the cost of very low throughput.

Propagation mechanisms

Free-space propagation assumes a direct line-of-sight path and produces a received-power relationship described by the Friis transmission equation:

Pr = Pt Gt Gr (λ/4πd)^2,

where Pt is transmit power, Gt and Gr are antenna gains, λ is wavelength, and d is distance. In decibels, free-space path loss increases by approximately 20 log10(d) and 20 log10(f), where f is frequency. The equation is an ideal reference, not a complete urban or indoor model.

Reflection occurs when a wave encounters a surface whose electrical dimensions are large relative to the wavelength. Diffraction allows energy to bend around edges and obstacles, while scattering results from rough surfaces or objects comparable to the wavelength. These mechanisms create multiple propagation paths. When replicas arrive with different phases, they can combine constructively or destructively, producing multipath fading.

Large-scale fading represents average power variation caused by distance and shadowing. A commonly used log-distance model is

PL(d) = PL(d0) + 10n log10(d/d0) + Xσ,

where n is the path-loss exponent and Xσ is a random shadowing term. Small-scale fading describes rapid fluctuations caused by multipath. Rayleigh fading is often used when no dominant direct path exists; Rician fading is more suitable when a strong line-of-sight component is present.

Propagation delay, Doppler shift, and coherence properties also matter. Relative motion changes the observed frequency by the Doppler shift and can make channel estimates outdated. Delay spread can cause intersymbol interference when a symbol duration is comparable to the spread of multipath arrivals. Orthogonal frequency-division multiplexing (OFDM) addresses this problem by dividing a broadband channel into many narrowband subcarriers and using a cyclic prefix.

Noise and interference

Thermal noise power is approximated by

N = kTB,

where k is Boltzmann's constant, T is absolute temperature, and B is receiver bandwidth. In decibels, receiver noise is commonly represented by noise figure (NF), which indicates degradation from the ideal thermal-noise limit. Increasing bandwidth can increase both capacity and collected noise, so bandwidth is not automatically beneficial.

Interference is energy from other transmitters or systems that affects reception. It may be co-channel, adjacent-channel, intersymbol, intermodulation, or self-interference. In dense IoT environments, interference management may be more important than raw transmit power. Techniques include channel planning, filtering, directional antennas, time coordination, power control, spreading, coding, listen-before-talk, and spatial reuse.

Modulation, Coding, and Multiplexing

Modulation maps digital symbols onto a carrier. In amplitude-shift keying, information changes amplitude; in frequency-shift keying, it changes frequency; and in phase-shift keying, it changes phase. Quadrature amplitude modulation (QAM) changes both amplitude and phase and can carry multiple bits per symbol, but higher-order QAM requires greater SNR and is more sensitive to distortion.

Spread-spectrum methods deliberately distribute a signal over a bandwidth larger than the minimum required. Direct-sequence spread spectrum multiplies data by a high-rate code, while frequency hopping changes among channels. Spreading can improve robustness against narrowband interference and support multiple users, although it generally reduces instantaneous spectral efficiency.

Orthogonal frequency-division multiplexing divides a channel into orthogonal subcarriers. It is effective against frequency-selective fading and supports adaptive modulation on individual subcarriers. Its disadvantages include high peak-to-average power ratio, sensitivity to frequency offset, and waveform complexity.

Coding adds controlled redundancy so that a receiver can detect or correct errors. Block codes operate on groups of bits, convolutional codes use memory, and modern systems use turbo, low-density parity-check, or polar codes. The coding rate is the ratio of information bits to transmitted bits. Lower coding rates improve reliability but consume more resources. Automatic repeat request (ARQ) retransmits failed packets, whereas hybrid ARQ combines retransmission with incremental error-correction information.

Multiplexing allows several signals to share a medium. Frequency-division multiplexing separates users by frequency, time-division multiplexing by time, and code-division multiplexing by spreading codes. Spatial multiplexing uses multiple antennas and independent propagation paths. Orthogonal multiple access methods allocate distinct resources, while non-orthogonal approaches permit controlled overlap followed by advanced detection. In IoT, the choice depends on device density, traffic regularity, synchronization capability, latency, and terminal complexity.

Link Budgets and Antenna Engineering

A link budget accounts for gains and losses between transmitter and receiver. A simplified decibel expression is

Pr(dBm) = Pt(dBm) + Gt(dBi) + Gr(dBi) - PL(dB) - Lc(dB) - Lm(dB),

where Lc includes cable and connector losses and Lm represents miscellaneous implementation losses. The link margin is

M = Pr - Sr,

where Sr is receiver sensitivity. A positive margin does not guarantee perfect service; engineers normally include a fade margin to accommodate shadowing, fading, aging, and environmental variation.

Receiver sensitivity depends on noise bandwidth, noise figure, required SINR, modulation, and coding. A simplified relation is

Sr = -174 + 10 log10(B) + NF + SNRrequired,

with values in dBm when bandwidth is in hertz. This equation demonstrates why narrowband operation can support long range: it reduces noise power and required received power, although it also limits data rate.

Antennas transform guided electrical energy into radiation and determine directionality, polarization, gain, efficiency, and impedance matching. Gain is not the same as power creation; it represents concentration of radiated energy in particular directions. An omnidirectional antenna is useful for mobile or randomly oriented devices, while a directional antenna can increase range and reduce interference when alignment is feasible. Polarization mismatch causes additional loss. Small IoT devices present difficult antenna problems because batteries, circuit boards, enclosures, human bodies, and nearby metal alter the radiation pattern.

Antenna design must therefore be validated in the assembled product, not only in simulation. Measurements should consider total radiated power, efficiency, radiation pattern, specific absorption constraints where relevant, and detuning across the intended installation environment.

Capacity, Reliability, and Resource Trade-offs

Shannon's idealized capacity relationship is

C = B log2(1 + SNR),

where C is channel capacity in bits per second and B is bandwidth. The theorem defines an upper bound under stated assumptions; it does not specify a particular modulation or guarantee practical performance. Increasing bandwidth produces approximately linear capacity growth at high SNR, while increasing power produces diminishing returns in the logarithmic term.

Spectral efficiency is throughput divided by occupied bandwidth. Energy efficiency is delivered information per unit energy, often measured in bits per joule. These objectives can conflict. A system using high-order modulation may achieve excellent spectral efficiency but require retransmissions and high processing power. A narrowband, low-rate system may achieve superior coverage and battery life but be unsuitable for firmware updates or real-time control.

Reliability must be specified operationally. For a sensor, a packet delivery probability of 99 percent may be inadequate if measurements are safety-critical, while a redundant application may tolerate occasional loss. Designers should distinguish average performance from tail performance, such as the fifth-percentile data rate or outage probability. Diversity—through time, frequency, coding, antenna, or route selection—reduces the likelihood that one fading event causes service failure.

Spectrum Engineering and Cognitive Radio

Radio spectrum is a finite physical resource but a reusable one. Allocation involves legal authorization, interference protection, technical conditions, and social priorities. Licensed spectrum gives an operator controlled rights, whereas unlicensed spectrum permits shared use subject to technical rules. Industrial, scientific, and medical bands support many IoT systems but may experience unpredictable contention.

Spectrum engineering includes band selection, channel bandwidth, transmit-power limits, emission masks, duty-cycle restrictions, coexistence rules, and geographic planning. A sound allocation policy considers propagation characteristics, incumbent users, device density, public safety, economic value, and enforcement capacity. Low-frequency bands generally offer better penetration and range, while higher frequencies can provide larger bandwidth but often suffer greater blockage and propagation loss.

Cognitive radio refers to systems capable of sensing or otherwise learning spectrum conditions and adapting parameters such as frequency, power, bandwidth, or access time. A cognitive cycle commonly includes spectrum sensing, opportunity identification, decision, transmission, and learning. Spectrum sensing may use energy detection, matched filtering, or cooperative sensing. Challenges include hidden incumbents, sensing uncertainty, false alarms, malicious signals, and the need to protect primary users.

Machine learning can assist channel selection, traffic prediction, anomaly detection, and radio-resource management. However, a learned policy must be constrained by regulation and safety requirements. Black-box decisions are problematic when they cause harmful interference or cannot demonstrate compliance. Data quality, distribution shift, adversarial manipulation, and explainability are important research concerns.

Energy-Aware Wireless Design

Energy is a central design variable in IoT. Total energy per useful bit includes sensing, processing, radio transmission, reception, idle listening, synchronization, security operations, and retransmissions. A radio that transmits at low power but remains awake continuously may consume more energy than a higher-power radio with an efficient sleep schedule.

Duty cycling reduces idle consumption by alternating between active and sleep states. Wake-up radios can reduce main-radio listening costs. Adaptive data rates, local aggregation, payload compression, edge processing, and event-triggered reporting reduce communications overhead. Power control should balance energy savings against increased packet loss and interference. Multi-hop routing can shorten individual links but adds relay overhead and creates dependence on intermediate nodes.

Energy harvesting from light, vibration, heat, or radio frequency fields can extend operational life, but harvested energy is variable. Energy-neutral design treats the device, workload, storage, and environment as a coupled system. It is inappropriate to claim perpetual operation without modeling seasonal availability, conversion efficiency, storage losses, and peak-power requirements.

An Illustrative Case Study: Agricultural Soil-Monitoring Network

Background

A research consortium plans to deploy soil-moisture, temperature, and salinity sensors across a 600-hectare agricultural region. Sensors are distributed irregularly, powered by small solar panels and batteries, and expected to operate for several years. Measurements are sent to gateways connected to a cloud analytics platform.

Situation and problem

The initial proposal uses a high-throughput short-range wireless technology because it is inexpensive and widely available. Field trials show that foliage, terrain variation, and gateway distance produce unstable links. Increasing transmit power improves some locations but rapidly reduces battery life and raises interference to neighboring installations.

Analysis

The engineering team first maps terrain, vegetation, gateway locations, and expected path-loss conditions. It measures packet delivery rather than relying only on received signal strength. A link budget is prepared using measured path-loss exponents, antenna gains, cable losses, receiver sensitivity, and a fade margin. The team identifies that the application requires only small periodic messages, not continuous high throughput. Consequently, sensitivity, coverage, and energy per delivered message are more important than peak data rate.

The team compares three approaches: additional gateways, a lower-rate long-range waveform, and a multi-hop mesh. Additional gateways improve reliability but increase installation and backhaul costs. The long-range option reduces infrastructure but has lower throughput and duty-cycle constraints. Mesh networking provides alternate paths but requires routing maintenance and increases dependence on intermediate nodes.

Approach and solution

The consortium selects a hybrid architecture. Long-range links carry routine measurements to strategically placed gateways, while local short-range links are used for maintenance and occasional high-volume data. Sensors use adaptive data rates, scheduled transmissions, acknowledgements for important messages, and sleep modes between measurement windows. Directional or elevated gateway antennas improve coverage without forcing sensors to use high power. The design also includes gateway redundancy in areas with difficult terrain.

The deployment is evaluated through a pilot covering representative soil types and elevations. Performance criteria include packet delivery, outage duration, energy consumption, latency of irrigation alerts, and maintenance effort. The data platform detects abnormal link behavior and distinguishes sensor failure from radio interference.

Outcome

The hybrid approach provides a better balance than simply increasing transmit power. Routine traffic is delivered with low energy expenditure, while high-value alerts receive stronger reliability treatment. The pilot also reveals that antenna placement and enclosure design are as important as the selected protocol. Seasonal vegetation changes are incorporated into the fade margin and maintenance plan.

Learning points

The case illustrates five principles. First, application requirements should determine the radio architecture. Second, a link budget must include realistic losses and fading. Third, energy is consumed by listening and retransmission as well as transmission. Fourth, field measurement is essential because theoretical propagation models are incomplete. Finally, spectrum and infrastructure decisions should be evaluated jointly rather than in isolation.

Challenges, Limitations, and Common Design Errors

Wireless models necessarily simplify reality. The Friis equation does not represent obstruction, multipath, or antenna installation effects. Average path-loss models can hide localized dead zones. Simulations may use ideal channel estimates or traffic assumptions that are not valid in deployment. Laboratory coexistence tests may fail to reproduce industrial interference.

Common errors include confusing antenna gain with transmit power, treating received-signal strength as a complete reliability measure, ignoring receiver noise figure, selecting bandwidth without considering noise and regulatory limits, and calculating range without a fade margin. Engineers also sometimes optimize peak data rate while neglecting battery life, latency distribution, retransmissions, or maintenance access.

Security creates additional constraints. Jamming, spoofing, replay, rogue access points, and denial-of-service attacks may target the physical or link layer. Encryption does not prevent radio-frequency interference. Secure boot, authenticated updates, key management, frequency agility, anomaly detection, and resilient fallback modes should be considered in system architecture.

Regulatory, Ethical, and Professional Considerations

Wireless systems operate within national and international regulatory frameworks. Designers must comply with frequency allocations, power limits, bandwidth restrictions, emission masks, certification requirements, and restrictions on harmful interference. Operating in an unlicensed band does not mean that a device has unrestricted access or guaranteed quality of service.

Spectrum decisions can have distributional consequences. Connectivity projects may prioritize urban commercial users while neglecting rural or remote communities. Sensor deployments can collect location and behavioral data without meaningful consent. Researchers and practitioners should apply data minimization, transparency, privacy protection, and proportionality. Environmental responsibility includes reducing electronic waste, designing for repair, selecting efficient protocols, and assessing the energy used by network infrastructure and cloud processing.

Professional practice requires reproducible measurement, explicit assumptions, uncertainty reporting, and separation of measured results from modeled predictions. Claims about range, battery life, or reliability should state the test conditions and confidence limits.

Emerging Directions

Current research is extending wireless systems in several directions. Reconfigurable intelligent surfaces aim to influence propagation using programmable electromagnetic structures, although practical deployment and control overhead remain open issues. Integrated sensing and communication seeks to use the same waveform and infrastructure for data transfer and environmental perception. Non-terrestrial networks are expanding coverage through satellites and high-altitude platforms, while private cellular systems provide managed connectivity for industrial sites.

Machine learning is increasingly used for channel estimation, beam selection, spectrum prediction, and energy management. Open radio-access architectures encourage modularity and programmable control, but they also expand the software attack surface. Terahertz and millimeter-wave systems offer large bandwidth but face severe blockage and hardware challenges. Research on semantic communication asks whether systems can transmit task-relevant meaning rather than complete raw data. These developments should be evaluated against measurable requirements, not novelty alone.

Best-Practice Design Workflow

A defensible wireless design normally proceeds as follows:

1. Define application traffic, reliability, latency, mobility, coverage, lifetime, and safety requirements.

2. Characterize the physical environment through maps, measurements, and propagation assumptions.

3. Select candidate spectrum bands and verify regulatory conditions.

4. Prepare link budgets with realistic antenna, feeder, noise, interference, and fade-margin terms.

5. Compare modulation, coding, access, and multiplexing options using throughput, reliability, and energy metrics.

6. Design antennas and validate them in the final enclosure and installation.

7. Model coexistence and test under representative interference and load.

8. Prototype, measure, and revise using statistical rather than single-point results.

9. Incorporate security, privacy, monitoring, firmware update, and maintenance mechanisms.

10. Document assumptions, uncertainties, compliance evidence, and operational limits.

Chapter Summary

Wireless IoT design combines electromagnetic propagation, information theory, signal processing, networking, regulation, and energy management. Propagation mechanisms produce path loss, shadowing, multipath, Doppler effects, and delay spread. Modulation and coding determine how information is represented and protected, while multiplexing and multiple access allow users to share spectrum. Link budgets connect transmit power, antenna gains, losses, receiver sensitivity, and fade margin to practical coverage.

Shannon capacity provides an upper bound that clarifies the relationship among bandwidth, SNR, and data rate, but practical systems must also address interference, reliability, latency, and energy. Spectrum engineering balances licensed and unlicensed access, coexistence, regulatory protection, and efficient reuse. Cognitive radios and machine learning can adapt radio resources, but their decisions must remain compliant, secure, and explainable. Effective IoT design is therefore an evidence-based process in which application requirements, field measurements, energy constraints, and governance considerations are treated as one system problem.

Key Terms

Propagation: The travel of electromagnetic energy through space and materials.

Path loss: Reduction in received signal power caused by distance and propagation conditions.

Shadowing: Large-scale signal variation caused by obstacles and environmental screening.

Multipath fading: Signal fluctuation caused by multiple paths combining at the receiver.

Doppler shift: Frequency change caused by relative motion between transmitter and receiver.

SNR: Ratio of desired signal power to noise power.

SINR: Ratio of desired signal power to interference plus noise power.

Modulation: Mapping information onto a carrier waveform.

Coding rate: Ratio of information bits to total coded bits.

OFDM: Multicarrier technique using orthogonal subcarriers to transmit broadband data.

Link budget: Accounting of gains and losses determining received power and link margin.

Receiver sensitivity: Minimum received power required for specified performance.

Fade margin: Additional link allowance for unpredictable signal degradation.

Antenna gain: Directional concentration of radiated power relative to a reference antenna.

Spectral efficiency: Data rate transmitted per unit bandwidth.

Energy efficiency: Useful information delivered per unit of energy.

Shannon capacity: Theoretical maximum reliable data rate for a specified bandwidth and SNR.

Cognitive radio: Radio system capable of sensing and adapting to spectrum conditions.

Duty cycling: Alternating active and sleep periods to reduce energy consumption.

Coexistence: Ability of multiple wireless systems to operate with acceptable mutual interference.

Review Questions

Short Answer Questions

1. What is the difference between path loss and multipath fading?

2. Why does increasing receiver bandwidth also increase thermal noise power?

3. Define link margin and fade margin.

4. What problem does OFDM address in a frequency-selective channel?

5. Why is spectral efficiency different from energy efficiency?

Long Answer Questions

1. Explain the major radio-wave propagation mechanisms and discuss their implications for IoT deployments.

2. Derive the logic of a wireless link budget and explain the engineering meaning of each major term.

3. Compare modulation, coding, and retransmission as methods for improving wireless reliability.

4. Discuss the relationship among Shannon capacity, bandwidth, SNR, spectral efficiency, and practical system constraints.

5. Explain how licensed, unlicensed, and cognitive-radio approaches address spectrum-sharing problems.

Analytical / Application Questions

1. An IoT device has limited battery capacity but must communicate through a heavily obstructed industrial site. Which changes to bandwidth, coding, antenna configuration, transmit schedule, and gateway placement would you investigate first, and why?

2. A proposed system reports excellent received signal strength but poor packet delivery during busy hours. Analyze possible causes and propose a measurement and mitigation plan.

3. Design an evaluation framework for comparing a long-range low-rate network with a short-range high-rate mesh for environmental monitoring. Specify performance metrics, test conditions, and decision criteria.

22. Multiple Choice Questions with Answers

Question 1

Which factor is directly represented by the fade margin in a wireless link budget?

A. The number of information bits per symbol

B. Allowance for unpredictable signal degradation

C. The legal ownership of a frequency band

D. The antenna’s physical size only

Correct Answer: B. Allowance for unpredictable signal degradation

Question 2

A system doubles its occupied bandwidth while transmit power and other conditions remain unchanged. What is the most immediate effect on thermal noise power?

A. It is approximately halved

B. It remains exactly zero

C. It approximately doubles

D. It becomes independent of temperature

Correct Answer: C. It approximately doubles

Question 3

Why can a narrowband IoT system achieve long range despite having low throughput?

A. Narrow bandwidth reduces collected noise power

B. Narrowband signals eliminate all interference

C. Narrowband antennas create power from nothing

D. Narrowband systems do not require receiver sensitivity

Correct Answer: A. Narrow bandwidth reduces collected noise power

Question 4

Which technique is most directly intended to combat intersymbol interference caused by multipath delay spread?

A. Increasing packet payload without coding

B. OFDM with an appropriate cyclic prefix

C. Removing the receiver filter

D. Using only amplitude changes

Correct Answer: B. OFDM with an appropriate cyclic prefix

Question 5

A receiver observes strong signal power but also strong co-channel interference. Which metric is most appropriate for evaluating the link?

A. Wavelength only

B. SINR

C. Antenna height only

D. Coding rate alone

Correct Answer: B. SINR

Question 6

What is the principal trade-off associated with higher-order QAM?

A. Higher data rate per symbol but greater SNR requirement

B. Lower spectral efficiency and no sensitivity to noise

C. Elimination of the need for antennas

D. Guaranteed operation under all fading conditions

Correct Answer: A. Higher data rate per symbol but greater SNR requirement

Question 7

In the agricultural monitoring case, why was simply increasing sensor transmit power not selected as the main solution?

A. It would reduce received power at the gateway

B. It could improve some links but increase energy use and interference

C. Transmit power has no relationship to link performance

D. It would make all antennas omnidirectional

Correct Answer: B. It could improve some links but increase energy use and interference

Question 8

Which activity is part of a cognitive-radio cycle?

A. Ignoring incumbent users

B. Sensing spectrum conditions and adapting radio parameters

C. Assigning every device the same permanent channel

D. Removing regulatory constraints through software

Correct Answer: B. Sensing spectrum conditions and adapting radio parameters

Question 9

Which statement best describes Shannon capacity?

A. A guaranteed throughput for every commercial radio

B. An ideal upper bound relating bandwidth and SNR to reliable rate

C. A measure of antenna polarization mismatch

D. The maximum battery lifetime of a sensor

Correct Answer: B. An ideal upper bound relating bandwidth and SNR to reliable rate

Question 10

Why should an IoT antenna be tested in its final enclosure?

A. The enclosure can alter impedance, efficiency, and radiation pattern

B. Antenna behavior is independent of nearby materials

C. Testing an enclosure removes the need for a link budget

D. The enclosure determines the coding rate automatically

Correct Answer: A. The enclosure can alter impedance, efficiency, and radiation pattern