Radio Propagation and Link Engineering

Introduction and Context

Wireless communication systems exchange information through an electromagnetic propagation medium that is neither fixed nor perfectly predictable. Unlike a guided cable, the radio channel is affected by distance, frequency, terrain, buildings, vegetation, atmospheric conditions, antenna placement, mobility, and interference from other transmitters. Radio propagation therefore forms the physical foundation on which coverage, data rate, reliability, frequency reuse, and network capacity depend.

Link engineering is the systematic process of determining whether a radio link can meet specified performance requirements. It combines propagation models, antenna characteristics, transmitter and receiver parameters, noise analysis, fading margins, interference constraints, and regulatory limits. A satisfactory received signal level alone does not guarantee communication: the signal-to-noise ratio (SNR), signal-to-interference-plus-noise ratio (SINR), modulation, coding, receiver sensitivity, and availability target must also be considered.

This chapter develops the principal propagation mechanisms and shows how they are incorporated into link budgets and coverage engineering. The emphasis is on models that are useful for postgraduate analysis while clearly identifying their assumptions and limitations.

Learning Objectives

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

1. Explain free-space propagation and derive the Friis transmission relationship.

2. Distinguish reflection, diffraction, scattering, shadowing, penetration loss, and multipath fading.

3. Select and interpret suitable path-loss and propagation measurement models.

4. Construct a radio link budget using consistent units and engineering assumptions.

5. Analyze antenna gain, effective aperture, polarization, noise, interference, and receiver sensitivity.

6. Estimate coverage and fading margins for practical wireless systems.

7. Critically evaluate propagation measurements, empirical models, and contemporary data-driven approaches.

1. Electromagnetic Foundations of Radio Propagation

A radio wave is an electromagnetic wave characterized by frequency f, wavelength λ, electric-field polarization, phase, and power density. In a homogeneous medium, wavelength is

λ = c/f,

where c is the wave velocity, approximately 3 × 10^8 m/s in free space. In a material medium, velocity is lower and depends on relative permittivity and permeability.

The propagation environment is commonly described in terms of the transmitter (Tx), receiver (Rx), separation distance d, carrier frequency, antenna heights, and surrounding objects. The channel may be:

• Line-of-sight (LOS), with a dominant unobstructed direct path;
• Non-line-of-sight (NLOS), where the direct path is blocked;
• Quasi-static, when the environment changes slowly relative to the observation interval; or
• Time-varying, when terminals, vehicles, people, or scatterers move.

A useful distinction is between large-scale and small-scale channel behavior. Large-scale effects include distance-dependent path loss and shadowing over tens or hundreds of wavelengths. Small-scale effects include rapid multipath fading caused by the constructive and destructive combination of waves arriving over slightly different paths.

Power density and effective aperture

For an isotropic transmitter radiating power Pt, the average power density at distance d in free space is

S = Pt/(4πd^2).

If the transmitting antenna has gain Gt in the direction of the receiver, the density becomes

S = Pt Gt/(4πd^2).

An antenna receives power according to its effective aperture Ae. The relationship between effective aperture and receive gain Gr is

Ae = Gr λ^2/(4π).

Thus, the received power is

Pr = Pt Gt Gr [λ/(4πd)]^2.

This is the Friis transmission equation. It assumes free-space propagation, far-field operation, matched polarization, matched impedance, and a direct unobstructed path. If polarization mismatch is represented by a factor Lp between 0 and 1, then Pr is multiplied by Lp. Additional losses may be included as multiplicative factors less than one.

The corresponding free-space path loss is

LFS = (4πd/λ)^2,

or in decibels,

LFS(dB) = 20 log10(4πd/λ).

Using d in kilometres and f in megahertz,

LFS(dB) = 32.44 + 20 log10(d_km) + 20 log10(f_MHz),

where the rounded constant is approximately 32.44. The formula is valid only when the units match the stated constant.

The free-space equation shows two important effects. First, received power decreases with the square of distance, corresponding to 20 dB loss for every tenfold increase in distance. Second, at fixed physical antenna gain, higher frequency increases free-space path loss. However, practical antennas and arrays often provide frequency-dependent gain, beamforming, and aperture effects, so system-level frequency comparisons require more than the path-loss term alone.

2. Antennas in Link Engineering

An antenna converts guided electrical energy into electromagnetic radiation and performs the reverse operation at reception. Important parameters include gain, directivity, efficiency, beamwidth, polarization, impedance, radiation pattern, and effective aperture.

Directivity describes concentration of radiation relative to an isotropic radiator. Gain includes efficiency:

G = ηD,

where η is total radiation efficiency and D is directivity. Gain is dimensionless but is commonly expressed in dBi relative to an isotropic radiator. Antenna gain in dBd is referenced to a half-wave dipole; conversion is approximately G_dBi = G_dBd + 2.15.

The effective isotropic radiated power is

EIRP(dBm) = Pt(dBm) + Gt(dBi) − Ltx(dB),

where Ltx includes feeder, connector, and other transmitter-side losses. Regulatory limits may specify maximum conducted power, EIRP, or equivalent radiated power, depending on the band and jurisdiction.

Polarization mismatch reduces received power. For linearly polarized antennas with relative polarization angle ψ, the ideal power coupling factor is

Lp = cos^2 ψ.

At 90 degrees, ideal linear polarization coupling is zero. Real systems may still receive some energy because of depolarization, reflections, and imperfect antenna polarization, but the ideal relationship emphasizes the importance of polarization alignment.

The far-field region is required for the usual gain and Friis assumptions. A commonly used criterion for an antenna with largest dimension D is

d_FF ≥ 2D^2/λ.

Near-field measurements and links involving electrically large arrays require more careful treatment because the angular field distribution and impedance behavior vary with distance.

3. Propagation Mechanisms

3.1 Reflection

Reflection occurs when a wave encounters a surface whose electrical dimensions are large relative to wavelength and whose electromagnetic properties differ from those of the surrounding medium. Ground, walls, building facades, vehicles, and water surfaces can produce significant reflected components.

The amplitude and phase of a reflected wave are governed by Fresnel reflection coefficients, which depend on incidence angle, polarization, and the complex relative permittivity and conductivity of the surface. A reflected wave may add constructively or destructively to the direct wave. Consequently, a small change in position can produce a large change in received power.

For a two-ray ground-reflection model, a direct and a reflected ray are considered. At sufficiently large distances and under idealized conditions, the received power may decrease approximately as d^-4 rather than d^-2. This asymptotic result is not universal: it depends on antenna heights, reflection coefficient, phase difference, terrain, and the existence of a dominant ground reflection.

3.2 Diffraction

Diffraction allows radio energy to reach regions behind edges and obstacles. It becomes important when an obstacle blocks the direct path but the wave can bend around it. The knife-edge diffraction model is a useful first approximation.

Let h be the obstacle height above the straight line joining the antennas, and let d1 and d2 be the distances from the obstacle to the transmitter and receiver. The Fresnel diffraction parameter is

ν = h √[2(d1 + d2)/(λ d1 d2)].

For a single knife edge, the excess diffraction loss is often approximated by

Ld ≈ 6.9 + 20 log10[√((ν − 0.1)^2 + 1) + ν − 0.1] dB,

for ν greater than approximately −0.7. The model is approximate and should not be interpreted as exact for rounded hills, thick buildings, multiple edges, or highly reflective environments.

The first Fresnel zone is an important clearance concept. For a path with total length d = d1 + d2, the radius of the nth Fresnel zone at an intermediate point is

r_n = √(n λ d1 d2/d).

Obstruction within the first Fresnel zone causes phase and amplitude changes. Many link-planning practices seek substantial clearance of the first zone rather than merely visual LOS.

3.3 Scattering

Scattering results when waves encounter objects that are small or irregular relative to wavelength, including foliage, rough surfaces, lamp posts, utility structures, and particles. Scattered energy arrives from many directions and can sustain communication in areas without a direct path, but it generally introduces loss and delay spread.

At high frequencies, objects that appear electrically small at lower frequencies may become significant scatterers. This is one reason why propagation behavior cannot be inferred from geometric appearance alone.

3.4 Multipath and fading

A received signal is the vector sum of components with different amplitudes, delays, and phases. A simplified baseband channel impulse response is

h(t, τ) = Σ_i a_i(t) exp[jφ_i(t)] δ(τ − τ_i(t)),

where ai, φi, and τi are the amplitude, phase, and delay of path i. The received signal may therefore experience amplitude fading, phase distortion, delay spread, and Doppler spreading.

When no dominant path exists, the envelope is often modeled by a Rayleigh distribution. When a strong direct or specular component exists, a Rician distribution is often used. These are statistical models, not universal laws. Their suitability must be tested against measurements.

The root-mean-square delay spread summarizes temporal dispersion. A signal with bandwidth substantially greater than the channel's coherence bandwidth may experience frequency-selective fading. Conversely, a sufficiently narrowband signal may experience approximately flat fading. Motion causes Doppler shifts; for a component arriving at angle θ relative to the direction of motion,

f_D = (v/λ) cos θ,

where v is terminal speed. The maximum Doppler frequency magnitude is approximately v/λ.

4. Large-Scale Path Loss and Shadowing Models

A general link model separates deterministic or average path loss from random shadowing and small-scale fading:

Pr(dB) = Pt(dB) + Gt(dB) + Gr(dB) − PL(dB) − Xσ(dB) − Lmisc(dB),

where Xσ is a zero-mean random shadowing term under the selected convention, and Lmisc includes penetration, feeder, polarization, and implementation losses.

Log-distance model

The log-distance path-loss model is

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

where d0 is a reference distance in the far field, n is the path-loss exponent, and Xσ is commonly modeled as a zero-mean Gaussian random variable in dB with standard deviation σ. Typical n values depend strongly on environment: free space has n = 2; obstructed indoor or urban environments may have substantially larger values, while waveguiding corridors can produce lower values over limited ranges.

The reference loss PL(d0) should be calculated or measured consistently. The model is not intended for extrapolation far beyond the range over which n and σ were estimated.

Multi-slope and empirical models

Real channels often require different exponents at different distances. A breakpoint may occur in a two-ray environment, and indoor channels may change behavior between rooms, corridors, and floors. Empirical macrocell models such as Okumura-Hata, COST-231 extensions, and standardized cellular models incorporate frequency, antenna height, distance, and environment class. Their use is valid only within their specified frequency, distance, antenna-height, and environment ranges.

Modern standards-based models may include street orientation, building height, clutter class, line-of-sight state, atmospheric effects, and statistical shadowing. Ray tracing can model site-specific geometry but depends on accurate maps and material parameters. Purely data-driven machine-learning models can improve prediction in a local area, but they may fail when deployed outside the training distribution and can be difficult to interpret physically.

5. Penetration, Shadowing, and Coverage

Shadowing is slow variation in average received power caused by buildings, terrain, foliage, and other obstructions. It is distinct from multipath fading because shadowing varies over larger spatial scales. A log-normal shadowing model is convenient because losses in decibels become additive and approximately Gaussian in many measurement settings.

Building penetration loss depends on frequency, wall and window materials, incidence angle, construction, floor level, and building layout. It should not be represented by one universal value. A practical model may use separate losses for external walls, internal partitions, windows, floors, and entry points, with stochastic variation around each mean.

Coverage is not simply the area where received power exceeds a threshold. A coverage criterion may require:

• Received signal strength above a minimum level;
• SINR above the modulation and coding requirement;
• A specified probability of service, such as 95% of locations;
• A specified time availability, such as 99.9%; and
• Acceptable throughput, latency, and outage performance.

If shadowing Xσ is Gaussian in dB with standard deviation σ, and the mean received power is μr, the probability that received power exceeds threshold γ is

P(Pr ≥ γ) = 1 − Φ[(γ − μr)/σ],

where Φ is the standard normal cumulative distribution function. This relationship permits a statistical coverage margin to be selected. It applies to the assumed Gaussian shadowing model and does not itself account for fast fading or interference.

6. Noise, Interference, and Receiver Sensitivity

Thermal noise power over bandwidth B is

N = kTB,

where k is Boltzmann's constant and T is absolute temperature. In dBm, at reference temperature 290 K,

N(dBm) ≈ −174 + 10 log10[B(Hz)].

Receiver noise figure NF increases the effective noise power:

N_receiver(dBm) = −174 + 10 log10[B] + NF.

Receiver sensitivity depends on the required SNR or SINR. For a required demodulator SNR γ_req,

S_min(dBm) = N_receiver(dBm) + γ_req(dB) + Limplementation(dB).

The required SNR depends on modulation, coding rate, packet-error target, channel estimation, and implementation. It is therefore not a universal constant for a particular data rate.

Interference is often more important than thermal noise in dense networks. The relevant quantity is

SINR = S/(N + I),

where S is desired received power, N is noise power, and I is aggregate interference power. Interference may be co-channel, adjacent-channel, intermodulation, or external. Directional antennas, frequency planning, power control, scheduling, polarization discrimination, and interference coordination are common mitigation tools.

A system can be noise-limited when additional transmit power substantially improves performance, or interference-limited when interference dominates and power increases provide little benefit. This distinction affects network design: increasing antenna height or adding spectrum may be more effective than increasing transmit power in an interference-limited network.

7. Link Budgets and Engineering Margins

A link budget expresses received power in additive decibel form:

Pr(dBm) = Pt(dBm) + Gt(dBi) + Gr(dBi) − LFS(dB) − Ltx(dB) − Lrx(dB) − Lother(dB).

For a non-free-space link, LFS is replaced or supplemented by the selected path-loss, diffraction, penetration, and shadowing terms. The link margin is

M(dB) = Pr − S_min.

A positive margin is necessary but not sufficient for a specified availability. Additional margin may be allocated for shadowing, rain attenuation at high microwave frequencies, foliage variation, equipment aging, and multipath fading. Margins should not be added twice: for example, a measured path-loss model may already include average building penetration or shadowing.

Worked example: terrestrial microwave link

Consider a 6 GHz point-to-point link of length 10 km. The transmitter output is 20 dBm, the transmitter feeder loss is 2 dB, the transmit antenna gain is 24 dBi, the receive antenna gain is 24 dBi, and the receive feeder loss is 2 dB. Assume 1 dB of additional implementation loss. Estimate free-space received power and link margin for a receiver sensitivity of −78 dBm.

The free-space loss is

LFS = 32.44 + 20 log10(10) + 20 log10(6000)

= 32.44 + 20 + 75.56

≈ 128.00 dB.

The received power is

Pr = 20 − 2 + 24 − 128 + 24 − 2 − 1

= −65 dBm approximately.

Therefore,

M = −65 − (−78) = 13 dB.

This is the nominal free-space margin. A practical design must determine whether this 13 dB is adequate for rain, diffraction, multipath, antenna misalignment, and the required annual availability. If a 10 dB fade margin is reserved for the propagation environment, approximately 3 dB remains for unmodeled losses; whether that is adequate depends on the link objective and climate.

8. Propagation Measurement and Model Validation

Propagation models should be selected and validated through measurement rather than applied mechanically. A measurement campaign should define frequency, bandwidth, antenna heights, polarization, route or grid, sampling interval, terminal speed, weather, and calibration procedure.

A typical workflow is:

1. Calibrate transmit power, receiver gain, cables, connectors, and antenna patterns.

2. Record received power, position, time, channel state, and relevant environmental metadata.

3. Separate large-scale variation from small-scale fading using spatial averaging or filtering appropriate to wavelength and application.

4. Estimate model parameters such as path-loss exponent n and shadowing standard deviation σ.

5. Divide data into training and independent validation sets.

6. Report mean error, root-mean-square error, percentile error, and outage prediction accuracy.

7. Examine residuals for spatial correlation, frequency dependence, and systematic bias.

For linear regression of measured path loss in dB against log10(d), the slope estimates 10n under the log-distance model. However, correlated shadowing violates the assumption of independent errors and can make uncertainty estimates overly optimistic. Repeated measurements, cross-validation, and confidence intervals are therefore important.

Channel sounders may estimate impulse response, power delay profile, Doppler spectrum, and spatial correlation. Wideband sounding provides richer information than narrowband received-power measurements, but it requires synchronization, calibration, dynamic range, and substantial data processing.

9. Illustrative Case: Urban Small-Cell Coverage

Background

Consider an illustrative deployment of a 3.5 GHz small cell intended to provide outdoor service along a 500 m urban street and indoor service on the first floor of nearby buildings. The operator requires 95% location coverage at a minimum SINR of 5 dB for a selected modulation and coding profile.

Problem

A free-space calculation predicts adequate power throughout the street, but initial field tests show strong variations near corners and much poorer indoor performance. Some locations have acceptable received power but fail the SINR requirement because of interference from an adjacent small cell.

Analysis

The street contains LOS segments, corner NLOS regions, parked vehicles, glass facades, and pedestrians. Reflection from buildings creates multipath, while corners produce diffraction loss. Indoor penetration adds variable wall and window loss. The correct design metric is SINR, not received power alone. Measurements reveal that the shadowing standard deviation is approximately 7 dB in the selected outdoor route, while indoor locations show larger variation.

Approach and solution

The planning team uses separate LOS and NLOS path-loss models, adds measured building penetration categories, and simulates interference using the actual antenna patterns. The antenna is tilted and azimuthally adjusted to reduce overshoot. A second low-power node is placed near the worst corner rather than simply increasing the original transmit power. Indoor service is improved through a dedicated indoor access point in the most heavily occupied building. The final design includes a margin based on the 95th-percentile location requirement and validates the prediction with an independent drive test.

Outcome

The revised design improves coverage because it addresses diffraction and shadowing, and it improves user performance because it reduces interference. The case demonstrates that a higher nominal received level does not necessarily produce better service. It also illustrates the value of site-specific validation and the danger of extrapolating a free-space model into an obstructed urban environment.

Learning points

1. Propagation mechanisms must be mapped to the physical environment.

2. Coverage and capacity are coupled through SINR.

3. Statistical margins should correspond to explicit availability or location-probability targets.

4. Network topology, antenna pattern, and interference coordination may be more effective than transmit-power increases.

10. Design Practice, Regulation, and Contemporary Developments

Good link engineering begins with a clear service specification: frequency, bandwidth, data rate, latency, availability, outage criterion, terminal distribution, and regulatory jurisdiction. Engineers should document every gain and loss, maintain unit consistency, distinguish measured from assumed values, and state the validity range of every model.

Regulatory compliance includes frequency authorization, maximum power or EIRP, occupied bandwidth, out-of-band emissions, electromagnetic exposure requirements, antenna installation rules, and coexistence obligations. Safety assessments must consider workers and the public, particularly for high-gain transmit antennas and rooftop installations. Privacy and governance also matter when drive testing collects location-linked user or device data; data minimization, consent where required, access control, and anonymization are appropriate safeguards.

Contemporary systems use massive MIMO, beamforming, millimetre-wave frequencies, reconfigurable surfaces, integrated sensing and communication, and non-terrestrial links. These technologies make propagation more directional and frequency-dependent. Beam alignment, blockage, human-body shadowing, atmospheric absorption, and rapid beam switching become central design issues at millimetre-wave and sub-THz frequencies. Digital twins and machine-learning-assisted propagation prediction can accelerate planning, but they should be treated as engineering models requiring validation, uncertainty quantification, and protection against biased or nonrepresentative training data.

Challenges and Limitations

No single propagation model is accurate in every environment. Free-space models are transparent but idealized. Empirical models are practical but domain-limited. Ray tracing is physically informative but computationally and materially demanding. Statistical models capture variability but may obscure causal mechanisms. Machine-learning models can interpolate complex relationships but may fail under distribution shift.

Further limitations include inaccurate maps, uncertain material properties, antenna installation errors, changing vegetation, moving vehicles, weather dependence, and insufficient measurement sampling. Link budgets also commonly fail when engineers mix dBi and dBd, confuse dBm with dBW, neglect feeder losses, double-count margins, or use receiver sensitivity without specifying bandwidth and required error performance.

The appropriate engineering response is not to seek a universally perfect formula. It is to combine a model hierarchy: analytical calculations for first-order feasibility, empirical or site-specific models for planning, measurements for calibration, and operational monitoring for continued assurance.

Chapter Summary

Radio propagation is governed by distance-dependent spreading and by interactions with terrain, surfaces, obstacles, buildings, vegetation, and atmospheric conditions. The Friis equation provides the fundamental free-space relationship, while reflection, diffraction, scattering, and multipath explain deviations from ideal behavior. Large-scale path loss and shadowing determine average coverage; small-scale fading and Doppler determine rapid channel variation.

A link budget combines transmit power, antenna gains, feeder losses, path loss, penetration, interference, and implementation margins. Receiver sensitivity follows from thermal noise, bandwidth, noise figure, and required SNR. Practical coverage must be defined probabilistically and, in modern networks, through SINR and service performance rather than received power alone.

Propagation models range from free-space and log-distance equations to empirical, ray-tracing, and machine-learning approaches. Each has assumptions and a validity domain. Measurement campaigns, calibration, independent validation, regulatory compliance, and explicit uncertainty analysis are essential. Effective link engineering is therefore both a quantitative discipline and an evidence-based design process.

Key Terms

Free-space path loss: Ideal propagation loss caused by spherical spreading in an unobstructed homogeneous medium.

Friis transmission equation: Relationship between transmitted and received power for a far-field free-space link.

Antenna gain: Directivity including antenna efficiency, usually expressed in dBi.

Effective aperture: Equivalent receiving area of an antenna, related to gain by Ae = Grλ²/(4π).

Reflection: Redirection of a wave from a surface with different electromagnetic properties.

Diffraction: Bending and spreading of waves around edges or obstacles.

Scattering: Redistribution of electromagnetic energy by small or irregular objects.

Multipath: Reception of multiple copies of a signal through paths with different delays and phases.

Shadowing: Large-scale random variation in mean received power due to obstructions and environmental features.

Path-loss exponent: Parameter n describing distance dependence in the log-distance model.

Penetration loss: Additional attenuation caused by transmission through walls, windows, floors, or other materials.

Fresnel zone: Region around a direct path within which obstruction affects phase and received power.

Doppler shift: Frequency shift caused by relative motion between transmitter, receiver, and propagation paths.

Noise figure: Measure of receiver degradation of input signal-to-noise ratio.

SINR: Ratio of desired signal power to the sum of noise and interference power.

Link budget: Accounting of all gains and losses in a radio link.

Fade margin: Excess received power above the minimum required level reserved for propagation variability.

EIRP: Effective isotropic radiated power, including transmitter power and antenna gain minus transmitter losses.

Coherence bandwidth: Frequency range over which channel responses remain strongly correlated.

Ray tracing: Site-specific propagation prediction based on modeled geometric paths and material interactions.

Review Questions

Short Answer Questions

1. State the assumptions underlying the Friis transmission equation.

2. Distinguish shadowing from small-scale multipath fading.

3. Define antenna effective aperture and give its relationship to antenna gain.

4. Why is SINR often a more useful coverage metric than received power?

5. What is the purpose of a fade margin in a link budget?

Long Answer Questions

1. Explain reflection, diffraction, and scattering, and compare their effects in urban wireless channels.

2. Derive the free-space path-loss expression from the Friis transmission equation and discuss unit conventions.

3. Explain the log-distance shadowing model, including the meaning and estimation of its parameters.

4. Describe a rigorous propagation measurement campaign and the methods used to validate a model.

5. Discuss the advantages and limitations of analytical, empirical, ray-tracing, and machine-learning propagation models.

Analytical / Application Questions

1. A 2.4 GHz link has 100 m separation, isotropic antennas, and 20 dBm transmit power. Calculate the free-space received power before implementation losses, and state the assumptions used.

2. A receiver has 10 MHz bandwidth, 6 dB noise figure, and requires 8 dB SNR. Calculate its approximate sensitivity before any additional implementation margin.

3. A link has a nominal margin of 16 dB, but measurements show 8 dB shadowing standard deviation. Using a Gaussian shadowing assumption, discuss whether the margin is likely to support a 95% location-availability target and identify additional information required for a final decision.

Multiple Choice Questions with Answers

Question 1

Which condition is required for direct use of the Friis transmission equation?

A. The link must be in the antenna far field with an unobstructed path.

B. The receiver must operate in the near field.

C. The channel must contain only diffuse scattering.

D. The antennas must have zero gain.

Correct Answer: A. The link must be in the antenna far field with an unobstructed path.

Question 2

What is the free-space path loss at 1 km and 1 GHz, approximately?

A. 32.44 dB

B. 72.44 dB

C. 92.44 dB

D. 122.44 dB

Correct Answer: C. 92.44 dB

Question 3

For ideal linearly polarized antennas separated by 90 degrees in polarization angle, the polarization coupling factor is

A. 1

B. 0.707

C. 0.5

D. 0

Correct Answer: D. 0

Question 4

In the log-distance model, the parameter n primarily describes

A. Receiver noise figure

B. Distance-dependent path-loss rate

C. Antenna polarization angle

D. Signal bandwidth

Correct Answer: B. Distance-dependent path-loss rate

Question 5

Which phenomenon is most directly associated with rapid amplitude variation over distances comparable to a wavelength?

A. Small-scale multipath fading

B. Regulatory emission control

C. Thermal equilibrium

D. Cable impedance matching

Correct Answer: A. Small-scale multipath fading

Question 6

At 290 K, the thermal noise power in a 1 MHz bandwidth is approximately

A. −174 dBm

B. −144 dBm

C. −114 dBm

D. −84 dBm

Correct Answer: C. −114 dBm

Question 7

A receiver has thermal noise of −104 dBm and a noise figure of 5 dB. If the required SNR is 10 dB, the approximate sensitivity before implementation margin is

A. −109 dBm

B. −99 dBm

C. −89 dBm

D. −79 dBm

Correct Answer: C. −89 dBm

Question 8

A link has transmitted power 15 dBm, total antenna gains of 30 dB, and total path and hardware losses of 120 dB. Its received power is

A. −75 dBm

B. −90 dBm

C. 105 dBm

D. 165 dBm

Correct Answer: A. −75 dBm

Question 9

Which metric includes desired signal, thermal noise, and interference?

A. EIRP

B. Effective aperture

C. SINR

D. Fresnel radius

Correct Answer: C. SINR

Question 10

Why is independent validation important when fitting a propagation model?

A. It guarantees zero measurement noise.

B. It tests performance on data not used to estimate the model.

C. It eliminates the need for calibration.

D. It makes all environments follow free-space propagation.

Correct Answer: B. It tests performance on data not used to estimate the model.