Constellation Scaling Error

Constellation scaling error originates from a mismatch in the amplitude gain between the in-phase (I) and quadrature (Q) signal paths. When the I/Q Gain Ratio deviates from unity, the constellation is stretched along one axis and compressed along the other, transforming a square QPSK constellation into a rectangle. This impairment is typically caused by component tolerances in amplifiers, mixers, or filters within the analog front-end.

The error manifests as a systematic deviation from the ideal symbol locations. Key geometric effects include:

• I/Q Constellation Ellipticity: Circular point clusters become elliptical.
• Axis Compression/Expansion: Symbol amplitude is altered on only one axis.
• Non-Uniform Symbol Distances: The Euclidean distance between adjacent symbols becomes asymmetric, degrading Modulation Error Ratio (MER) performance.

While both are components of I/Q Imbalance, scaling error is distinct from Quadrature Skew. Scaling error is a pure amplitude error affecting the magnitude of the I or Q component. Quadrature skew is a phase error where the angle between the I and Q axes deviates from 90 degrees. Combined, these two impairments cause the complex Constellation Warping that forms a unique I/Q Distortion Signature.

The specific I/Q gain ratio is a highly stable, manufacturing-dependent parameter. It forms a critical component of the I/Q Constellation Distortion Profile used for Physical Layer Authentication. Because the gain mismatch is determined by fixed analog component values, it provides a robust, unclonable hardware identifier that is independent of the transmitted data payload.

Scaling error is quantified by calculating the I/Q Gain Ratio from a captured I/Q Constellation Diagram. This is done by measuring the average amplitude of received symbols along the I-axis versus the Q-axis. A ratio of 1.0 indicates perfect balance. The error is often expressed in decibels (dB) as 20 * log10(Gain_I / Gain_Q). This metric is a key input feature for Deep Learning Signal Identification models.

Digital Adaptive I/Q Correction algorithms can estimate and compensate for scaling error. These techniques apply a corrective gain factor to one of the signal paths in the digital baseband to restore the constellation to its ideal square geometry. However, for fingerprinting purposes, this correction is intentionally omitted to preserve the unique I/Q Constellation Distortion Uniqueness for device authentication.