Spectral Mask

A spectral mask is not a single value but a piecewise function specifying maximum emission levels at various frequency offsets. The mask typically becomes more stringent (lower power) as the offset from the carrier increases. For example, a 3GPP LTE mask might specify -25 dBm/MHz at a 5 MHz offset, tightening to -36 dBm/MHz at 10 MHz. This graduated structure reflects the practical reality that filtering becomes more effective further from the carrier, while also protecting adjacent channel users from excessive interference.

Regulatory masks distinguish between two emission regions:

• Out-of-Band Emissions: Unwanted power immediately adjacent to the assigned channel, caused by modulation and transmitter nonlinearity. These are specified with tight, frequency-dependent limits.
• Spurious Emissions: Emissions far from the carrier, including harmonics and parasitic signals. These are governed by absolute, often less stringent limits. This distinction is critical because the physical mechanisms—and thus the mitigation techniques like filtering vs. DPD—differ fundamentally between the two domains.

Spectral mask compliance is measured with a specific measurement bandwidth (e.g., 30 kHz, 1 MHz) that defines the resolution of the spectrum analyzer. A mask limit expressed as -13 dBm/MHz means the total power integrated over a 1 MHz window must not exceed -13 dBm. Using a narrower measurement bandwidth can reveal spectral regrowth peaks that would be averaged out in a wider window. Standards bodies carefully specify both the limit and the measurement bandwidth to ensure consistent, reproducible compliance testing across different test equipment.

Crest Factor Reduction (CFR) and spectral mask compliance are in constant tension:

• Aggressive CFR reduces PAPR and improves PA efficiency but introduces nonlinear distortion that causes spectral regrowth into adjacent channels.
• Conservative CFR preserves spectral purity but forces the PA to operate at a large back-off, sacrificing efficiency. The engineering challenge is designing a CFR algorithm that maximizes PAPR reduction while ensuring the resulting spectrum—after PA nonlinearity and DPD—remains strictly within the regulatory mask. This often requires iterative co-simulation of CFR, DPD, and PA models.

ACLR is the primary metric for quantifying spectral mask compliance in adjacent channels. It measures the ratio of transmitted power in the assigned channel to the power leaking into an adjacent channel, typically expressed in dBc. For example, 3GPP requires ACLR better than -45 dBc for LTE base stations. ACLR is directly degraded by PA nonlinearity, and the spectral mask sets the absolute power limits that translate into required ACLR values. DPD systems are explicitly designed to improve ACLR by 15-25 dB to meet these stringent mask requirements.

Each wireless standard defines its own spectral mask tailored to its channel bandwidth and coexistence requirements:

• LTE 20 MHz: Specifies limits at offsets of ±10 MHz, ±15 MHz, and ±20 MHz from the carrier center.
• 5G NR: Defines masks for flexible numerologies with subcarrier spacings from 15 kHz to 120 kHz, accommodating channel bandwidths up to 100 MHz (FR1) and 400 MHz (FR2).
• Wi-Fi 6: Uses a transmit spectrum mask with 0 dBr at the channel edge, dropping to -20 dBr at ±11 MHz and -45 dBr beyond ±30 MHz. Designers must verify compliance against the specific mask for their target standard.

The primary quantitative metric used to verify spectral mask compliance. ACLR measures the ratio of transmitted power within the assigned channel to power leaking into adjacent frequency channels.

• Measurement: Integrated power in the adjacent channel vs. in-band channel
• Typical requirements: -45 dBc for 4G LTE, -45 dBc for 5G NR (varies by band)
• Relationship to mask: ACLR is a single-number summary; the spectral mask is the continuous limit line
• Failure mode: Nonlinear power amplifier operation is the dominant cause of ACLR degradation

The phenomenon where nonlinear amplification causes a signal's spectrum to broaden beyond its original bandwidth, generating unwanted power in adjacent channels. Spectral regrowth is the physical mechanism that causes spectral mask violations.

• Root cause: Third-order intermodulation products from amplifier compression
• Key characteristic: Regrowth appears as a widening of the spectral shoulders
• Mitigation: Digital predistortion (DPD) linearizes the amplifier to suppress regrowth
• Measurement: Visible on a spectrum analyzer as elevated power in adjacent channels

Regulatory requirements defined by standards bodies (3GPP, ETSI, FCC) that specify the maximum permissible radiated power outside the licensed operating band. The spectral mask is the graphical representation of these limits.

• 3GPP TS 38.104: Defines masks for 5G NR base stations
• Frequency dependence: Limits become more stringent farther from the carrier
• Measurement bandwidth: Specified per frequency offset (e.g., 30 kHz, 1 MHz)
• Conformance testing: Requires calibrated spectrum analyzers and specific test modes

A signal conditioning technique that reduces the peak-to-average power ratio (PAPR) of a transmit waveform before amplification. CFR directly impacts spectral mask compliance by reducing the amplifier's exposure to high-peak excursions that drive nonlinearity.

• Trade-off: Aggressive CFR improves efficiency but introduces in-band distortion (EVM)
• Methods: Peak windowing, peak cancellation, and hard/soft clipping
• Spectral impact: Poorly designed CFR creates its own out-of-band emissions
• Joint optimization: CFR and DPD are often co-designed to balance EVM and ACLR

A measure of in-band signal quality that quantifies the deviation of received constellation points from their ideal positions. EVM and spectral mask compliance represent competing constraints in transmitter design.

• Regulatory limits: 3GPP specifies maximum EVM per modulation scheme (e.g., 3.5% for 64QAM)
• CFR trade-off: Reducing PAPR via clipping increases EVM
• DPD benefit: Linearization improves both EVM and ACLR simultaneously
• Measurement: Performed with vector signal analyzers on demodulated symbols

The intentional reduction of input drive level to operate a power amplifier in its linear region. Back-off is the most direct—but least efficient—method to ensure spectral mask compliance.

• Relationship to PAPR: Required back-off equals the signal's PAPR plus margin
• Efficiency penalty: A 10 dB back-off can reduce PA efficiency from 50% to below 20%
• DPD alternative: Digital predistortion allows operation closer to saturation (reduced back-off)
• Thermal implications: Higher back-off reduces dissipated heat but wastes DC power