Introduction to Radio Frequency measurements for Cellular and Wireless Communication Systems

While the modulation and demodulation techniques described in the text books is perfect, in reality they are far from perfect. Due to the tolerance of different components used in manufacturing of Radio Frequency (RF) chipsets, we notice different impairments in the transmitted and received RF signals. In this presentation, we will introduce and discuss the RF measurements for Cellular and Wireless Communication Systems

What this presentation is about and why it matters

This talk is an accessible introduction to practical radio-frequency (RF) measurement for cellular and wireless systems. It connects textbook modulation ideas (baseband, upconversion, single- and double-sideband) to the messy realities of hardware: mismatched I/Q channels, carrier leakage, oscillator frequency error and phase noise, DAC/ADC clock jitter, and nonlinearity from amplifiers. The presenter explains how those impairments show up in spectra and constellations, why they increase Error Vector Magnitude (EVM), and how engineers use measurements to find and fix problems during prototype and production testing.

Why it matters: modern wireless systems rely heavily on DSP and precise RF front-ends. If designers ignore real-world impairments, a theoretically correct modem will fail in production or in the field. Understanding what to measure, how impairments manifest, and how test workflows (golden units, conducted vs OTA testing, DMRS/PTRS reference signals) are used to diagnose issues is essential for anyone shipping radios at scale.

Who will benefit the most from this presentation

• DSP engineers who build algorithms for demodulation, calibration, and impairment estimation.
• RF/analog engineers and system architects who need to understand how component tolerances and PCB/layout choices affect DSP metrics.
• Test and validation engineers responsible for manufacturing test flows and test-fixture design (conducted vs. over-the-air).
• Students and researchers learning how abstract modulation theory connects to real hardware and measurement practice.

What you need to know

Before watching, refreshing these basic ideas will help you follow the examples and Q&A:

• Baseband vs RF: A baseband signal occupies frequencies around 0 Hz. Upconversion multiplies baseband I/Q by carrier sinusoids to place the signal around a carrier frequency $f_c$.
• I/Q modulation: Typical quadrature upconversion uses I(t) and Q(t) with a 90° phase offset: for an ideal modulator you can write $s_{RF}(t)=I(t)\cos(2\pi f_c t)-Q(t)\sin(2\pi f_c t)$. Imbalances in gain or phase between I and Q create mirror-image spectral components.
• Sidebands: Double-Sideband Suppressed-Carrier (DSB-SC) places both upper and lower sidebands around $f_c$. Single-Sideband (SSB) transmits only one sideband to save bandwidth and power but requires better hardware/synchronization.
• EVM (Error Vector Magnitude): A practical scalar metric of modulation quality. If $s_{ref}$ is the ideal symbol and $s_{rx}$ the received symbol, the error vector is $e=s_{rx}-s_{ref}$ and a common definition is \(\mathrm{EVM}=\frac{\sqrt{E[|e|^2]}}{\sqrt{E[|s_{ref}|^2]}}\) (reported as a ratio or dB/%). EVM integrates the effects of many impairments.
• Carrier frequency/phase errors: A frequency offset makes constellations rotate over time; phase noise spreads subcarrier energy (important in OFDM systems).
• DAC/ADC impairments: Clock jitter, sampling rate offset and quantization noise can raise EVM and cause symbol timing errors.
• Power amplifier nonlinearity: Limited dynamic range produces intermodulation products and spectral regrowth (in-band and out-of-band emissions).
• Testing basics: Conducted testing (cabled) is common for sub-6 GHz; millimeter-wave devices often require over-the-air (OTA) chambers and beamforming-aware fixtures.
• Reference signals: Embedded pilots (e.g., DMRS, PTRS in 5G) are used to estimate the channel and help separate device impairments from the propagation channel.

Glossary

• I/Q imbalance — mismatch in amplitude or phase between the in-phase (I) and quadrature (Q) paths, causing mirror-image (image) interference.
• Carrier leakage (LO leakage) — unwanted carrier tone present in the transmitted spectrum due to imperfect cancellation or leakage paths.
• Phase noise — random, small rapid variations in oscillator phase that broaden spectral lines and degrade OFDM subcarriers.
• Frequency offset — a static or slowly varying error between transmitter and receiver carrier frequencies; results in constellation rotation.
• EVM (Error Vector Magnitude) — a normalized measure of modulation error; combines effects from noise, imbalance, offset and distortion.
• DSB-SC / SSB — double-sideband suppressed-carrier vs single-sideband modulation; SSB saves bandwidth but is harder to implement.
• Conducted vs OTA testing — direct cable connection (conducted) vs over-the-air measurements inside shielded chambers (OTA), the latter used for millimeter-wave and antenna testing.
• DMRS / PTRS — demodulation and phase-tracking reference signals in 5G used for channel and impairment estimation.
• Intermodulation distortion (IMD) — nonlinear mixing of signals in an amplifier producing unwanted spurs at sum/difference frequencies.
• Golden device — a reference unit with known-good performance used to validate the test setup and separate DUT errors from tester errors.

A few final words about the talk

This presentation is a clear, practical bridge between DSP theory and the messy realities of real-world radios. The speaker mixes concise conceptual explanations with measured examples and manufacturing context (prototype cycles, golden units, and test throughput constraints). If you work on modulation, calibration, test automation, or RF system design, you’ll find the talk motivating: it highlights where DSP matters most in production and gives a realistic picture of the engineering tradeoffs between cost, performance and test time.

Enjoy the talk — it’s approachable, grounded in real products, and a good starting point for deeper study of impairment estimation algorithms and production test strategies.

This overview is AI-generated from the session transcript. Spot an issue? Let us know.