Inferensys

Glossary

In-Phase & Quadrature (IQ) Components

The two orthogonal carrier signal components where the in-phase (I) component is modulated by a cosine wave and the quadrature (Q) component is modulated by a sine wave, forming the basis for representing any signal state in the complex plane.
Stylish WeWork-like workspace with hot desks and document wall, professional searching through enterprise knowledge base on a mounted ultrawide display, warm industrial pendants overhead.
COMPLEX SIGNAL REPRESENTATION

What is In-Phase & Quadrature (IQ) Components?

The two orthogonal carrier signal components where the in-phase (I) component is modulated by a cosine wave and the quadrature (Q) component is modulated by a sine wave, forming the basis for representing any signal state in the complex plane.

In-Phase (I) and Quadrature (Q) components are the two orthogonal basis signals that decompose a modulated carrier into a complex-valued representation, where the I component modulates a cosine carrier and the Q component modulates a sine carrier of the same frequency. This decomposition allows any bandpass signal to be equivalently represented by a complex baseband envelope, mapping directly to the x-axis and y-axis of a constellation diagram.

The orthogonality of the I and Q carriers—derived from the 90-degree phase shift between cosine and sine—enables two independent data streams to be transmitted simultaneously over the same bandwidth without mutual interference. In a direct-conversion receiver, the incoming signal is mixed with a local oscillator to recover the I component and with a 90-degree phase-shifted version of that oscillator to recover the Q component, reconstructing the transmitted symbol's position in the complex plane for minimum distance decoding.

ORTHOGONAL SIGNAL DECOMPOSITION

Key Characteristics of IQ Representation

The in-phase (I) and quadrature (Q) components form the orthogonal basis functions that enable any digitally modulated signal to be represented as a single complex-valued sample, mapping directly to a point in the constellation diagram.

01

Orthogonal Basis Functions

The I and Q components are modulated onto cosine and sine carriers respectively, which are orthogonal over a symbol period. This orthogonality means the two data streams do not interfere with each other, effectively doubling the spectral efficiency by transmitting two independent real-valued signals simultaneously within the same bandwidth. The mathematical inner product of a cosine and sine wave over a complete cycle is zero, ensuring perfect separability at the receiver.

02

Complex Envelope Representation

Any bandpass signal can be expressed as the real part of a complex baseband signal multiplied by a carrier. The complex baseband signal is precisely I(t) + jQ(t), where j is the imaginary unit. This representation shifts the analysis from high-frequency carrier waveforms to low-frequency complex phasors, dramatically simplifying digital signal processing, simulation, and the generation of constellation diagrams.

03

Instantaneous Amplitude and Phase

The Cartesian I and Q coordinates map directly to polar coordinates:

  • Instantaneous Amplitude (Envelope): √(I² + Q²). Constant for pure PSK, varying for QAM.
  • Instantaneous Phase: arctan(Q/I). The primary carrier attribute modulated in PSK schemes. This dual representation is fundamental to understanding how QAM simultaneously modulates both amplitude and phase to create dense constellation grids.
04

IQ Imbalance Artifact

In a perfect system, the I and Q branches have exactly equal gain and a 90-degree phase offset. IQ imbalance occurs when the physical receiver hardware deviates from this ideal. Gain mismatch causes the received constellation to stretch into a rectangular shape, while phase error skews it into a parallelogram. This hardware impairment creates an image of the signal spectrum, degrading the error vector magnitude (EVM) and classification accuracy.

05

Baseband Sampling Theorem

Because the IQ decomposition shifts the signal's bandwidth from a passband centered at the carrier frequency down to baseband (centered at 0 Hz), the required Nyquist sampling rate is determined by the signal's bandwidth, not its maximum frequency. A 20 MHz wide signal at a 2.4 GHz carrier only requires a complex sampling rate of >20 MHz, not >4.8 GHz. This makes modern digital receivers and software-defined radio practical.

06

Rotating Phasor Visualization

A constant unmodulated carrier is a static point on the IQ plane. Phase modulation rotates this point around the origin, while amplitude modulation moves it radially. A continuous phase shift, such as that caused by a carrier frequency offset (CFO) , manifests as a constant-speed rotation of the entire constellation. This visual diagnostic allows engineers to instantly identify synchronization errors in a scatter plot.

IQ COMPONENTS

Frequently Asked Questions

Clear, technically precise answers to the most common questions about the in-phase and quadrature components that form the foundation of modern digital modulation and signal representation.

In-phase (I) and quadrature (Q) components are the two orthogonal basis signals that represent any modulated carrier as a single complex number. The I component is the projection of the signal onto a cosine carrier wave (0° reference), while the Q component is the projection onto a sine carrier wave (90° offset). Together, they form a complex envelope s(t) = I(t) + jQ(t), where the instantaneous amplitude is √(I² + Q²) and the instantaneous phase is arctan(Q/I). This quadrature representation is the universal language of software-defined radio, enabling arbitrary phase and amplitude modulation without requiring separate hardware paths for every possible signal state. The orthogonality of sine and cosine ensures that I and Q channels do not interfere with each other, effectively doubling the spectral efficiency compared to single-dimensional modulation.

Prasad Kumkar

About the author

Prasad Kumkar

CEO & MD, Inference Systems

Prasad Kumkar is the CEO & MD of Inference Systems and writes about AI systems architecture, LLM infrastructure, model serving, evaluation, and production deployment. Over 5+ years, he has worked across computer vision models, L5 autonomous vehicle systems, and LLM research, with a focus on taking complex AI ideas into real-world engineering systems.

His work and writing cover AI systems, large language models, AI agents, multimodal systems, autonomous systems, inference optimization, RAG, evaluation, and production AI engineering.