I/Q Imbalance

The amplitude mismatch between the I and Q branches, defined as the ratio or difference in gain. This causes the constellation to stretch along one axis, compressing or expanding the other.

• Measured in dB or as a percentage deviation from unity
• Results in elliptical constellation distortion
• Produces an image signal proportional to the gain error magnitude
• Typically specified as < 0.1 dB for high-performance modulators

The deviation from the ideal 90-degree phase offset between I and Q local oscillator signals. This causes inter-symbol interference and constellation rotation.

• Measured in degrees of deviation from orthogonality
• Causes skewed constellation points that rotate toward each other
• Generates an image signal with phase-dependent amplitude
• Typical specifications: < 1 degree for precision modulators

An unwanted constant voltage added to the baseband I or Q signal, typically from local oscillator self-mixing or component mismatch. This manifests as carrier leak at the center of the transmitted spectrum.

• Produces a spurious tone at the carrier frequency
• Degrades spectral mask compliance and wastes transmit power
• Can be static (fixed offset) or dynamic (varying with temperature)
• Corrected through DC offset cancellation loops

A type of imbalance where gain and phase errors vary across the signal bandwidth, caused by mismatched anti-aliasing filters, trace lengths, or component parasitics.

• Requires complex FIR filter correction rather than scalar compensation
• Manifests as frequency-selective image response
• Critical for wideband signals (>100 MHz) in 5G NR systems
• Modeled using widely-linear system representations

The relative timing delay between sampling clocks or data paths of the I and Q channels. This is a form of frequency-dependent imbalance causing linear phase distortion across the signal bandwidth.

• Measured in picoseconds or fractions of a sample period
• Produces frequency-dependent constellation rotation
• Becomes significant at high sample rates (>1 GSPS)
• Compensated using fractional delay filters or all-pass networks

The primary metric quantifying a system's ability to suppress the unwanted image signal generated by I/Q imbalance. Expressed as the power ratio between the desired signal and its image in dB.

• Calculated from gain and phase error: IRR ≈ -10 log₁₀((ε² + φ²)/4)
• Typical uncorrected IRR: 25-35 dB
• With digital compensation: >60 dB achievable
• Directly impacts adjacent channel leakage ratio (ACLR)

The amplitude mismatch component of I/Q imbalance. It is defined as the ratio or difference in gain between the I and Q branches. This causes the ideal square constellation to stretch along one axis, compressing the other, which directly degrades Error Vector Magnitude (EVM). Unlike phase error, gain imbalance is a scalar error that can often be corrected with a simple multiplication factor in the digital baseband.

The deviation from the ideal 90-degree phase offset between the I and Q local oscillator signals. This non-orthogonality causes inter-symbol interference and a characteristic rotation and skewing of the constellation. It is a primary contributor to the generation of an image signal that falls directly on top of the desired spectrum, making it impossible to filter out in the analog domain.

The key performance metric for quantifying I/Q imbalance severity. IRR is the power ratio between the desired signal and the unwanted image signal, expressed in decibels (dB). A high IRR indicates excellent balance. Typical targets for high-order modulation like 256-QAM require IRRs exceeding 40 dB, which is unachievable in analog silicon without digital compensation.

A complex mismatch where gain and phase errors vary across the signal bandwidth. This is caused by physical mismatches in analog components like anti-aliasing filters, trace lengths, or DAC responses. Correcting this requires a complex FIR filter (an adaptive I/Q equalizer) rather than a simple scalar, as the image is no longer a simple mirror but a dispersed, filtered version of the signal.

A digital linearization technique where the baseband I and Q signals are intentionally distorted with an inverse model of the modulator's imbalance. By applying a widely-linear transformation matrix before the DAC, the pre-distorted signal cancels out the analog impairments, resulting in a clean, orthogonal output at the antenna. This is often combined with Digital Pre-Distortion (DPD) for the power amplifier.

An advanced signal processing technique that extracts imbalance parameters directly from the statistical properties of the modulated signal, such as its circularity (properness). Unlike pilot-based methods, it requires no dedicated training sequence, allowing for real-time, in-service tracking of time-varying impairments like thermal drift without sacrificing spectral efficiency.