RF Engineering Foundations and System Metrics

Introduction and Context

Radio-frequency (RF) engineering concerns the generation, transmission, propagation, processing, and reception of electromagnetic signals at frequencies where circuit dimensions become electrically significant. Although “RF” is often used broadly for frequencies from a few kilohertz to hundreds of gigahertz, practical RF circuit design commonly addresses frequencies from approximately 3 kHz to 300 GHz, corresponding to wavelengths from 100 km to 1 mm. Microwave engineering generally emphasizes the upper portion of this range, while millimetre-wave engineering addresses frequencies above about 30 GHz.

At low frequencies, circuit interconnections can often be treated as ideal wires. At RF, the same interconnection may behave as a transmission line, an antenna, a resonator, or a source of unwanted coupling. Resistive loss, dielectric loss, radiation, parasitic capacitance and inductance, impedance mismatch, noise, nonlinear distortion, and electromagnetic compatibility therefore become central design concerns.

An RF system is normally analyzed as a chain of functional blocks: source, oscillator, modulator, power amplifier, filter, antenna, propagation channel, receiving antenna, low-noise amplifier, mixer, intermediate-frequency stages, demodulator, and baseband processor. System metrics such as gain, noise figure, bandwidth, linearity, sensitivity, dynamic range, and error-vector magnitude connect the behavior of individual circuits to overall communication performance.

Learning Objectives

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

1. Explain fundamental RF terminology and classify frequency bands.

2. Relate frequency, wavelength, phase, propagation velocity, and impedance.

3. Use decibel and dBm representations correctly, including cascaded gain and loss.

4. Explain impedance matching and calculate power transfer for common cases.

5. Distinguish occupied bandwidth, noise bandwidth, channel bandwidth, and sampling bandwidth.

6. Model thermal noise and calculate noise power, noise figure, and receiver sensitivity.

7. Explain compression, intermodulation, intercept points, dynamic range, and adjacent-channel interference.

8. Construct and interpret transmitter, receiver, and link budgets.

9. Evaluate practical trade-offs among gain, noise, linearity, power consumption, and bandwidth.

1. RF Terminology and Frequency Classification

A radio-frequency signal is an electrical or electromagnetic signal whose frequency is sufficiently high that propagation and distributed effects matter. The carrier frequency, denoted by \(f_c\), is the nominal frequency around which information is conveyed. A signal may occupy a band around this carrier because of modulation.

The wavelength is the spatial period of a sinusoidal wave. In a homogeneous medium,

\[

\lambda = \frac{v_p}{f}

\]

where \(\lambda\) is wavelength in metres, \(v_p\) is phase velocity in metres per second, and \(f\) is frequency in hertz. In free space, \(v_p=c\approx 3.00\times10^8\) m/s. In a nonmagnetic dielectric with relative permittivity \(\epsilon_r\), an approximate nondispersive relation is \(v_p\approx c/\sqrt{\epsilon_r}\). This approximation does not fully describe dispersive transmission lines or materials with frequency-dependent properties.

The International Telecommunication Union (ITU) commonly uses the following broad frequency designations:

| Band | Frequency range |

|---|---:|

| VLF | 3–30 kHz |

| LF | 30–300 kHz |

| MF | 300 kHz–3 MHz |

| HF | 3–30 MHz |

| VHF | 30–300 MHz |

| UHF | 300 MHz–3 GHz |

| SHF | 3–30 GHz |

| EHF | 30–300 GHz |

These labels describe frequency ranges, not specific services. Actual spectrum allocation is governed nationally and internationally. Allocations may be exclusive, shared, licensed, licence-exempt, primary, or secondary. Cellular, satellite, radar, broadcasting, aeronautical, scientific, and industrial applications may occupy different portions of the spectrum. A designer must therefore distinguish the physical frequency range from the regulatory permission to transmit.

Important RF terms include:

• Baseband: the information-bearing signal before translation to a carrier frequency.
• Passband: the frequency range surrounding a carrier after modulation.
• Radio channel: an assigned frequency range and associated technical conditions for communication.
• Tuning: adjustment of a circuit to select or generate a desired frequency.
• Harmonic: a frequency equal to an integer multiple of a fundamental frequency.
• Spurious emission: an unwanted emission outside the intended signal spectrum.
• In-band distortion: distortion that falls within the wanted channel.
• Out-of-band emission: energy outside the allocated or intended bandwidth.
• Electromagnetic compatibility (EMC): the ability of equipment to operate satisfactorily without causing or suffering unacceptable interference.

2. Wave Behavior and Distributed Circuits

2.1 Sinusoidal waves and phase

A sinusoidal voltage may be written as

\[

v(t)=V_m\cos(2\pi ft+\phi)

\]

where \(V_m\) is peak amplitude, \(f\) is frequency, and \(\phi\) is phase relative to a reference. The angular frequency is \(\omega=2\pi f\). For a travelling wave, the phase also varies with position:

\[

v(z,t)=V_m\cos(\omega t-\beta z+\phi)

\]

where \(\beta=2\pi/\lambda\) is the phase constant in radians per metre. The negative sign represents propagation in the positive \(z\)-direction.

A circuit can be treated as lumped when its largest physical dimension is much smaller than the signal wavelength. A common engineering guideline is that lumped approximation is increasingly questionable when the dimension exceeds approximately \(\lambda/10\), although the exact limit depends on allowable phase error and circuit sensitivity. At RF, even a short printed-circuit trace may require transmission-line treatment.

2.2 Transmission lines

A transmission line is characterized by distributed inductance and capacitance, together with distributed resistance and conductance. Its characteristic impedance, for a low-loss line, is approximately

\[

Z_0\approx\sqrt{\frac{L'}{C'}}

\]

where \(L'\) and \(C'\) are inductance and capacitance per unit length. For a general line,

\[

Z_0=\sqrt{\frac{R'+j\omega L'}{G'+j\omega C'}}.

\]

Here \(R'\), \(L'\), \(G'\), and \(C'\) are distributed parameters per unit length. The general expression applies to sinusoidal steady state under the usual uniform-line assumptions.

When a line is terminated by a load \(Z_L\) different from \(Z_0\), part of the incident wave is reflected. The voltage reflection coefficient is

\[

\Gamma=\frac{Z_L-Z_0}{Z_L+Z_0}.

\]

The voltage standing-wave ratio (VSWR), for \(|\Gamma|<1\), is

\[

\mathrm{VSWR}=\frac{1+|\Gamma|}{1-|\Gamma|}.

\]

A matched load has \(Z_L=Z_0\), so \(\Gamma=0\) and VSWR = 1. Mismatch causes reduced delivered power, ripple, frequency-dependent response, and potentially excessive voltage or current at standing-wave maxima.

For a source with Thevenin voltage \(V_{th}\) and source impedance \(Z_s\), maximum average power is delivered to a purely resistive load when \(R_L=R_s\). For complex impedances, the maximum-power condition is conjugate matching:

\[

Z_L=Z_s^*.

\]

This condition maximizes power at a single frequency under linear steady-state assumptions. It does not always maximize system performance: a low-noise amplifier may require a noise-optimum source impedance rather than a power-match impedance, and a power amplifier may require a load selected for efficiency or linearity.

2.3 S-parameters

At RF and microwave frequencies, voltage and current measurements are difficult to define directly at ports because the measurement apparatus itself perturbs the circuit. Scattering parameters, or S-parameters, describe travelling-wave relationships. For a two-port network,

\[

\begin{bmatrix}b_1\\b_2\end{bmatrix}

=\begin{bmatrix}S_{11}&S_{12}\\S_{21}&S_{22}\end{bmatrix}

\begin{bmatrix}a_1\\a_2\end{bmatrix}

\]

where \(a_i\) and \(b_i\) are incident and reflected wave amplitudes at port \(i\), defined with respect to reference impedances. \(S_{11}\) is input reflection, \(S_{22}\) is output reflection, \(S_{21}\) is forward transmission, and \(S_{12}\) is reverse transmission. S-parameters are usually measured with a vector network analyzer and are valid for specified frequency, bias, temperature, and reference impedance conditions.

3. Decibels, Power, and Voltage Ratios

The decibel is a logarithmic representation of a ratio. For powers \(P_1\) and \(P_2\),

\[

G_{dB}=10\log_{10}\left(\frac{P_2}{P_1}\right).

\]

For equal impedances, the corresponding voltage or current ratio is

\[

G_{dB}=20\log_{10}\left(\frac{V_2}{V_1}\right)

=20\log_{10}\left(\frac{I_2}{I_1}\right).

\]

The voltage form must not be used without considering impedance. In general, power is \(P=V_{rms}^2/R\) for a resistive impedance, so different impedances require explicit accounting.

An absolute power level is often expressed in dBm:

\[

P_{dBm}=10\log_{10}\left(\frac{P}{1\ \mathrm{mW}}\right).

\]

Thus 0 dBm = 1 mW, 10 dBm = 10 mW, 20 dBm = 100 mW, and 30 dBm = 1 W. Conversion back to watts is

\[

P[\mathrm{W}]=10^{(P_{dBm}-30)/10}.

\]

For a 50-ohm system, the RMS voltage associated with a power \(P\) is \(V_{rms}=\sqrt{P R}\). At 0 dBm, \(P=1\) mW, giving approximately 0.224 V RMS in 50 ohms.

Cascaded gains and losses add in decibels. If a signal passes through gains \(G_1,G_2\) and losses \(L_1,L_2\), all expressed in dB with losses represented as negative gains, then

\[

G_{total,dB}=G_1+G_2-L_1-L_2.

\]

This additive property is one reason dB is indispensable in RF link budgets.

4. Bandwidth and Frequency Selectivity

Bandwidth is the frequency span over which a specified condition is satisfied. The condition must always be stated. A filter may have a 3-dB bandwidth, an insertion-loss bandwidth, a group-delay bandwidth, or a regulatory occupied bandwidth.

For a simple resonator, the quality factor is

\[

Q=\frac{f_0}{B_{3dB}}

\]

where \(f_0\) is resonant frequency and \(B_{3dB}\) is the 3-dB bandwidth, provided the response is adequately represented by a single resonant mode. High \(Q\) implies narrow bandwidth and often high selectivity, but it may also increase sensitivity to component tolerances and reduce tuning range.

The occupied bandwidth is the frequency range containing a specified percentage of total signal power, often 99% or a value defined by a standard. The channel bandwidth is the assigned frequency interval, which may exceed occupied bandwidth to provide guard bands. The equivalent noise bandwidth is the bandwidth of an ideal rectangular filter that passes the same integrated white-noise power as the actual filter. It is generally not identical to the 3-dB bandwidth.

Time and bandwidth are related by modulation and filtering constraints. Narrow filtering improves noise and interference rejection but can cause intersymbol interference, waveform distortion, or loss of modulation sidebands. Wide bandwidth preserves modulation fidelity but admits more noise and interference.

5. Noise in RF Systems

Noise is an unwanted random fluctuation that limits detectability and information capacity. Important sources include thermal noise, shot noise, flicker noise, avalanche noise, oscillator phase noise, and externally generated atmospheric, man-made, and cosmic noise.

5.1 Thermal noise

For a resistor at absolute temperature \(T\), the available thermal-noise power over bandwidth \(B\) is

\[

N=kTB

\]

where \(k=1.380649\times10^{-23}\) J/K is Boltzmann’s constant. This available-power relation assumes a matched source and receiver and white noise over the relevant bandwidth. At room temperature, the thermal noise density is approximately -174 dBm/Hz at 290 K.

The mean-square open-circuit noise voltage of a resistor is

\[

\overline{v_n^2}=4kTRB

\]

for resistance \(R\), bandwidth \(B\), and conditions under which the resistor model is valid. The voltage is a random quantity with zero mean in the ideal model; the equation gives its mean-square value.

5.2 Noise figure and noise factor

The noise factor of a network is

\[

F=\frac{(S/N)_{in}}{(S/N)_{out}},

\]

where signal-to-noise ratios are measured under defined source and reference conditions. Noise figure is

\[

NF=10\log_{10}F\quad\text{dB}.

\]

An ideal noiseless network has \(F=1\) and NF = 0 dB. Real passive loss at reference temperature has a noise factor approximately equal to its linear loss, so an attenuator before a low-noise amplifier can seriously degrade receiver sensitivity.

For cascaded stages, Friis’ formula is

\[

F_{total}=F_1+\frac{F_2-1}{G_1}+\frac{F_3-1}{G_1G_2}+\cdots

\]

where all \(F_i\) and gains \(G_i\) are linear power quantities, not dB values. The formula assumes a matched cascade and consistent reference temperatures. It shows why the first active stage usually dominates receiver noise performance: its gain suppresses the input-referred noise contribution of later stages.

5.3 Sensitivity and signal-to-noise ratio

For a receiver with noise figure \(NF\) in dB and noise bandwidth \(B\), the input noise power is approximately

\[

N_{dBm}\approx -174+10\log_{10}(B)+NF

\]

at 290 K, with \(B\) in hertz. If the required output or demodulator signal-to-noise ratio is \((S/N)_{req}\) in dB, an approximate minimum input signal, or sensitivity, is

\[

S_{min,dBm}=N_{dBm}+(S/N)_{req}.

\]

This is an engineering estimate. Real sensitivity may additionally depend on implementation loss, coding gain, fading margin, phase noise, interference, detector threshold, and bit-error-rate requirements.

6. Linearity and Large-Signal Performance

An RF circuit is linear when superposition holds. Practical amplifiers are linear only over a finite signal range. As input power increases, gain compression, harmonic distortion, and intermodulation products appear.

The 1-dB compression point is the input or output power at which the actual gain is 1 dB below the extrapolated small-signal gain. It is commonly denoted \(P_{1dB,in}\) or \(P_{1dB,out}\). The input-referred and output-referred values differ by approximately the small-signal gain, subject to the convention used.

For a two-tone input at frequencies \(f_1\) and \(f_2\), third-order intermodulation products occur at

\[

2f_1-f_2\quad\text{and}\quad 2f_2-f_1.

\]

Because these products can fall close to the desired tones, third-order distortion is particularly important in receivers and transmitters. The third-order intercept point (IP3) is an extrapolated intersection of the fundamental and third-order product output lines. It is not normally a physically reached operating point. Input and output intercept points are related approximately by

\[

OIP3\approx IIP3+G

\]

when all quantities are in dBm and \(G\) is small-signal gain in dB.

For a two-tone test in the linear extrapolation region, if the fundamental output is \(P_{out}\) dBm and the third-order product is \(P_{IM3}\) dBm, then

\[

OIP3\approx P_{out}+\frac{P_{out}-P_{IM3}}{2}.

\]

The approximation depends on equal-tone test conditions and the usual cubic model.

6.1 Dynamic range

Dynamic range is the ratio between the largest tolerable signal and the smallest usable signal. It must be defined with respect to constraints. A receiver’s spurious-free dynamic range (SFDR) is often limited at the high end by third-order intermodulation and at the low end by the noise floor. For an idealized two-tone system, a commonly used estimate is

\[

SFDR_{dB}\approx \frac{2}{3}(IIP3-N_{in})

\]

where both \(IIP3\) and input noise power \(N_{in}\) are in dBm. This expression assumes third-order products, equal tone powers, and a specified comparison convention. It should not be treated as universal when blockers, reciprocal mixing, quantization noise, or detection thresholds dominate.

A high-linearity receiver is not automatically a low-noise receiver. Increasing bias current may improve IP3 while degrading power consumption; adding attenuation may improve overload tolerance while worsening sensitivity. RF design is therefore a constrained optimization problem rather than a single-metric maximization task.

7. Link Budgets and System-Level Performance

A link budget accounts for gains and losses from transmitter output to receiver input. In dB form,

\[

P_r=P_t+G_t+G_r-L_{path}-L_{misc}

\]

where powers and gains are in dBm or dB, \(P_t\) is transmitted power, \(G_t\) and \(G_r\) are antenna gains, and losses include path, polarization, feeder, atmospheric, implementation, and miscellaneous losses.

For free-space propagation, the Friis transmission equation is

\[

P_r=P_tG_tG_r\left(\frac{\lambda}{4\pi R}\right)^2

\]

where gains are linear relative to isotropic radiators, \(R\) is separation, and the antennas are in the far field with polarization alignment and line-of-sight conditions. The corresponding free-space path loss is

\[

L_{FS,dB}=20\log_{10}\left(\frac{4\pi R}{\lambda}\right).

\]

A practical path may require additional terms for shadowing, multipath fading, rain attenuation, atmospheric absorption, cable losses, and antenna mismatch. A fade margin is extra link power budget reserved to meet availability requirements during adverse propagation.

The link budget must be connected to receiver sensitivity. A link is viable when received signal power exceeds sensitivity by the required margin:

\[

M=P_r-S_{min}.

\]

The margin is not necessarily a guarantee of quality because interference, fading statistics, modulation-specific thresholds, and time availability requirements must also be considered.

8. Transmitter and Receiver Performance Metrics

8.1 Transmitter metrics

Key transmitter metrics include:

• Output power: delivered RF power under specified load, frequency, modulation, and duty cycle.
• Power-added efficiency (PAE): \(PAE=(P_{out}-P_{in})/P_{DC}\), where powers are consistent and the amplifier is operated under defined conditions.
• Gain: ratio of output to input power, often specified as small-signal or large-signal gain.
• Adjacent-channel leakage ratio (ACLR): power leaked into an adjacent channel relative to power in the assigned channel.
• Error-vector magnitude (EVM): normalized distance between measured and ideal modulation symbols, usually expressed as a percentage or dB.
• Phase noise: spectral density of short-term oscillator phase fluctuations, often expressed in dBc/Hz at an offset from the carrier.
• Spurious and harmonic suppression: compliance with emission limits.
• Load-pull performance: dependence of output power, efficiency, and linearity on complex load impedance.

A transmitter designed only for maximum saturated power may fail spectral-mask or EVM requirements. Linear modulation usually requires output back-off, which reduces efficiency. Digital predistortion can compensate some amplifier nonlinearities but increases system complexity and depends on model accuracy, bandwidth, feedback, and memory effects.

8.2 Receiver metrics

Important receiver metrics include:

• Gain and gain flatness: amplification and its frequency variation.
• Noise figure: degradation of signal-to-noise ratio.
• Sensitivity: minimum usable input signal under a stated performance criterion.
• Selectivity: ability to reject signals outside the wanted channel.
• Blocking: degradation caused by a strong off-channel signal.
• Desensitization: reduction in wanted-signal performance due to interference or overload.
• Linearity: resistance to compression and intermodulation.
• Image rejection: suppression of the unwanted mixing frequency in a superheterodyne receiver.
• Reciprocal mixing: conversion of strong off-channel noise through oscillator phase noise into the desired channel.
• Frequency accuracy and stability: ability to remain tuned to the assigned channel.

The receiver architecture affects these metrics. A superheterodyne receiver provides filtering and gain at an intermediate frequency, but image response and oscillator leakage must be managed. A direct-conversion receiver avoids an intermediate frequency but may suffer from DC offset, flicker noise, even-order distortion, I/Q imbalance, and local-oscillator leakage. A low-IF architecture trades some of these issues against image-rejection requirements. Direct-RF sampling can reduce analog mixing but transfers stringent requirements to the data converter, clock, front-end filter, and digital processing.

9. Worked Example: Cascaded Receiver Sensitivity and Linearity

Consider an illustrative 2.4-GHz receiver with a 1-MHz noise bandwidth. Its first stage is a low-noise amplifier with gain 15 dB and noise figure 2 dB. A mixer follows with conversion loss 6 dB and noise figure 8 dB. Assume matched stages and a 290-K reference.

Convert the first-stage quantities to linear form:

\[

G_1=10^{15/10}=31.62,

\qquad F_1=10^{2/10}=1.585.

\]

The mixer has gain \(G_2=10^{-6/10}=0.251\) and noise factor \(F_2=10^{8/10}=6.31\). Friis’ formula gives

\[

F_{total}=1.585+\frac{6.31-1}{31.62}

=1.585+0.168\approx1.753.

\]

Therefore,

\[

NF_{total}=10\log_{10}(1.753)\approx2.44\ \mathrm{dB}.

\]

The thermal noise in 1 MHz is

\[

-174+10\log_{10}(10^6)=-174+60=-114\ \mathrm{dBm}.

\]

Including the receiver noise figure gives an input noise power of approximately

\[

N_{in}=-114+2.44=-111.56\ \mathrm{dBm}.

\]

If the demodulator requires 10 dB signal-to-noise ratio, the approximate sensitivity is

\[

S_{min}=-111.56+10=-101.56\ \mathrm{dBm}.

\]

This result is conditional: it excludes implementation loss, interference, fading, coding gain, and any difference between equivalent noise bandwidth and the stated 1 MHz bandwidth.

Suppose the receiver has an input third-order intercept point of -10 dBm. Using the idealized SFDR estimate,

\[

SFDR\approx\frac{2}{3}[-10-(-111.56)]

=\frac{2}{3}(101.56)\approx67.7\ \mathrm{dB}.

\]

This is an estimate for the stated model, not a complete specification. A real receiver may be limited by a blocker, phase noise, ADC range, or a different intermodulation mechanism.

10. Illustrative Case: Urban 2.4-GHz Sensor Link

Background

A postgraduate design team is developing a short-range industrial sensor link at 2.4 GHz. The transmitter delivers 10 dBm to a 2-dBi antenna. The receiver uses a 2-dBi antenna and has a sensitivity of -95 dBm for the selected modulation and packet-error-rate target. The nominal separation is 100 m.

Problem

The team must determine whether the link has adequate nominal margin and whether regulatory and engineering constraints have been considered. The environment includes metal machinery, possible multipath, and interference from nearby unlicensed devices.

Analysis

At 2.4 GHz, the wavelength is approximately

\[

\lambda=\frac{3\times10^8}{2.4\times10^9}=0.125\ \mathrm{m}.

\]

The free-space path loss at 100 m is

\[

L_{FS}=20\log_{10}\left(\frac{4\pi(100)}{0.125}\right)

\approx80.05\ \mathrm{dB}.

\]

The nominal received power is therefore

\[

P_r=10+2+2-80.05\approx-66.05\ \mathrm{dBm}.

\]

The nominal sensitivity margin is

\[

M=-66.05-(-95)=28.95\ \mathrm{dB}.

\]

Approach and solution

The team should not treat the 29-dB result as guaranteed performance. It should subtract feeder and connector losses, estimate polarization and body losses, characterize antenna installation, and include a fading margin based on measured or modeled channel statistics. A spectrum survey should identify co-channel and adjacent-channel interference. Receiver blocking and front-end compression should be tested with strong interferers. The transmitter must meet regional power, bandwidth, spectral-mask, and duty-cycle rules.

A practical validation plan includes conducted measurements, radiated chamber measurements, over-the-air packet tests, temperature and supply-voltage tests, and operation near representative machinery. Adaptive channel selection or retransmission may improve reliability, but these features increase latency and protocol complexity.

Outcome and learning points

The free-space calculation indicates a strong nominal link, but real reliability depends on fading and interference rather than path loss alone. Link budgets are decision tools, not substitutes for measurement. The case also illustrates that antenna gains, noise performance, regulatory compliance, receiver blocking, and system protocol behavior must be evaluated together.

11. Design Practice, Limitations, and Governance

Good RF design begins with a complete specification: frequency range, modulation, channel spacing, output power, noise bandwidth, sensitivity criterion, linearity targets, temperature range, supply conditions, impedance environment, and applicable regulations. Measurements must state calibration plane, reference impedance, detector type, resolution bandwidth, averaging method, and uncertainty.

Common design errors include adding dBm values rather than converting to linear power when appropriate, applying Friis’ formula to dB quantities, confusing 3-dB bandwidth with noise bandwidth, ignoring impedance in voltage-ratio calculations, using ideal free-space loss in obstructed environments, and quoting noise figure without source impedance or temperature conditions.

Regulatory compliance is part of engineering responsibility. RF equipment must respect spectrum allocations, emission masks, human-exposure limits, electromagnetic compatibility rules, and certification requirements. Deliberate interference, unapproved emissions, and unsafe exposure are both technical and ethical failures. Engineers should document assumptions, preserve measurement traceability, protect user and network privacy, and avoid claiming performance beyond validated conditions.

Contemporary RF systems increasingly combine analog RF, mixed-signal conversion, digital calibration, machine-learning-assisted predistortion, beamforming, and software-defined radio. AI may assist in spectrum sensing, calibration, fault detection, and adaptive resource allocation, but it does not remove the need for physical models, test evidence, explainable limits, and regulatory oversight. Data-driven algorithms can fail under distribution shift, adversarial interference, or unrepresented hardware aging.

Chapter Summary

RF engineering treats electrical signals as electromagnetic waves whose propagation, impedance, bandwidth, noise, and nonlinear behavior determine system performance. Frequency and wavelength establish when distributed effects become important. Transmission lines, reflection coefficient, VSWR, and S-parameters provide practical tools for analyzing RF interconnections and networks.

Decibels simplify cascaded gain and loss calculations, while dBm expresses absolute power relative to 1 mW. Correct use requires attention to impedance and to the distinction between linear and logarithmic quantities. Bandwidth must be defined by a specific criterion, and equivalent noise bandwidth may differ from 3-dB bandwidth.

Thermal noise follows \(kTB\), and noise figure quantifies signal-to-noise degradation. Friis’ formula demonstrates the importance of the first receiver stage. Sensitivity depends on noise bandwidth, noise figure, and required signal-to-noise ratio, but practical performance also includes interference, fading, implementation loss, and modulation requirements.

Large-signal metrics—compression, IP3, intermodulation, blocking, and dynamic range—are essential because a receiver or transmitter must accommodate both weak desired signals and strong unwanted signals. Link budgets combine transmitter power, antenna gains, path loss, and miscellaneous losses to estimate received power and margin. Finally, RF design is a system-level optimization involving noise, linearity, efficiency, selectivity, cost, power consumption, measurement uncertainty, and regulation.

Key Terms

RF signal: An electrical or electromagnetic signal in a frequency range where RF propagation and distributed effects are relevant.

Carrier frequency: Nominal frequency around which a modulated signal is centered.

Wavelength: Spatial period of a wave, equal to phase velocity divided by frequency.

Transmission line: Distributed circuit used to convey high-frequency signals.

Characteristic impedance: Impedance ratio associated with a travelling wave on a uniform line.

Reflection coefficient: Ratio of reflected to incident wave amplitude at a discontinuity.

VSWR: Voltage standing-wave ratio caused by an incident and reflected wave.

S-parameter: Scattering parameter describing incident and outgoing travelling waves at a network port.

Decibel: Logarithmic measure of a power or, under stated impedance conditions, voltage ratio.

dBm: Absolute power level referenced to 1 mW.

Bandwidth: Frequency interval satisfying a specified response or power criterion.

Equivalent noise bandwidth: Rectangular bandwidth passing the same white-noise power as a real filter.

Thermal noise: Random noise generated by thermal agitation of charge carriers.

Noise figure: Noise-factor value expressed in decibels.

Sensitivity: Minimum input signal meeting a specified receiver performance criterion.

1-dB compression point: Signal level at which gain falls 1 dB below its small-signal extrapolation.

IP3: Extrapolated third-order intercept point used to characterize cubic nonlinearity.

Dynamic range: Range between the smallest usable and largest tolerable signal under defined constraints.

Link budget: Accounting of transmitted power, gains, losses, and received power.

EVM: Normalized error between measured and ideal modulation symbols.

Review Questions

Short Answer Questions

1. Why can an RF printed-circuit trace need to be treated as a transmission line?

2. Define reflection coefficient and state the condition for a matched load.

3. What is the difference between dB and dBm?

4. State the thermal-noise power relation and define its variables.

5. Why does the first stage of a receiver strongly influence cascaded noise figure?

Long Answer Questions

1. Explain the physical meaning of wavelength, phase constant, characteristic impedance, and VSWR in a transmission-line system.

2. Derive or explain the use of Friis’ cascaded noise-factor formula and discuss its assumptions.

3. Compare 1-dB compression, IP3, and SFDR as measures of RF linearity.

4. Explain how bandwidth affects noise, selectivity, modulation fidelity, and receiver sensitivity.

5. Discuss the principal performance metrics of an RF transmitter and receiver and the trade-offs among them.

Analytical / Application Questions

1. A receiver has a 200-kHz noise bandwidth, 4-dB noise figure, and requires 12 dB SNR. Estimate its sensitivity at 290 K. State the assumptions.

2. A two-stage cascade has a first-stage gain of 20 dB and noise figure of 2 dB, followed by a 10-dB-loss passive filter at 290 K. Calculate the approximate total noise figure in dB using linear quantities.

3. Design a preliminary link budget for a 5.8-GHz, 1-km line-of-sight link. Specify reasonable antenna gains and miscellaneous losses, calculate free-space path loss, and determine the required transmitter power for a stated received-power target. Justify every assumption.

Multiple Choice Questions with Answers

Question 1

For a 50-ohm system, what power level corresponds to 0 dBm?

A. 0 W

B. 1 mW

C. 10 mW

D. 1 W

Correct Answer: B. 1 mW

Question 2

A load is perfectly matched to a transmission line when which condition holds?

A. ZL = 0

B. ZL = infinity

C. ZL = Z0

D. ZL = -Z0

Correct Answer: C. ZL = Z0

Question 3

What is the approximate free-space wavelength at 2.4 GHz?

A. 1.25 m

B. 0.125 m

C. 12.5 m

D. 0.0125 m

Correct Answer: B. 0.125 m

Question 4

Which quantity is calculated by kTB for a matched source over bandwidth B?

A. Thermal-noise power

B. Third-order intercept point

C. Antenna gain

D. Reflection coefficient

Correct Answer: A. Thermal-noise power

Question 5

In Friis’ noise formula, why is the first receiver stage especially important?

A. Its gain suppresses the contribution of later stages when referred to the input

B. It always has zero noise figure

C. It determines the carrier frequency

D. It eliminates all interference

Correct Answer: A. Its gain suppresses the contribution of later stages when referred to the input

Question 6

A power gain of 20 dB corresponds to what linear power ratio?

A. 2

B. 10

C. 20

D. 100

Correct Answer: D. 100

Question 7

Which products are third-order intermodulation products for two tones f1 and f2?

A. f1 + f2 and f1 - f2

B. 2f1 and 2f2

C. 2f1 - f2 and 2f2 - f1

D. 3f1 and 3f2 only

Correct Answer: C. 2f1 - f2 and 2f2 - f1

Question 8

The 1-dB compression point indicates that the actual gain has fallen by what amount from its small-signal extrapolation?

A. 0.1 dB

B. 1 dB

C. 3 dB

D. 10 dB

Correct Answer: B. 1 dB

Question 9

Which statement about the free-space Friis transmission equation is correct?

A. It assumes far-field line-of-sight propagation under specified polarization conditions

B. It applies only to heavily obstructed urban channels

C. It excludes antenna gains

D. It is independent of wavelength

Correct Answer: A. It assumes far-field line-of-sight propagation under specified polarization conditions

Question 10

A receiver has a noise bandwidth of 1 MHz and a noise figure of 3 dB. Its approximate input noise power at 290 K is:

A. -174 dBm

B. -171 dBm

C. -114 dBm

D. -111 dBm

Correct Answer: D. -111 dBm