What Is QAM and How Constellation Maps to Binary

Quadrature Amplitude Modulation (QAM) is an efficient modulation technique that transmits digital data by altering both the amplitude and phase of carrier signals. This article explains the fundamentals of QAM, the structure of its constellation diagram, and the exact demodulation process used to interpret incoming analog symbols and map them back into binary bitstreams.

What Is Quadrature Amplitude Modulation (QAM)?

QAM is a hybrid modulation scheme that transmits two independent carrier waves—an In-phase (\(I\)) component and a Quadrature (\(Q\)) component—on the same frequency. The \(Q\) carrier is phase-shifted by 90 degrees relative to the \(I\) carrier.

By independently modulating the amplitude of both the \(I\) and \(Q\) carriers and summing them together, QAM creates unique combinations of amplitude and phase. Each unique state represents a “symbol,” allowing multiple binary bits to be transmitted within a single signal change. The total number of states determines the QAM order. For example, 16-QAM represents 4 bits per symbol (\(2^4 = 16\)), while 64-QAM represents 6 bits per symbol (\(2^6 = 64\)).

The Constellation Diagram

A constellation diagram is a two-dimensional geometric representation of all possible symbols a QAM system can transmit.

Each point on the grid represents an ideal symbol state characterized by a specific coordinate \((I, Q)\), which corresponds to a distinct binary word.

How Constellation Points Map Back to the Binary Number System

When a signal arrives at a receiver, noise, interference, and channel distortion cause the received coordinates to deviate from their ideal constellation locations. The receiver converts these physical signals back into digital binary numbers through the following structured process:

1. Coherent Demodulation (Extracting \(I\) and \(Q\))

The receiver multiplies the incoming analog signal by local reference oscillators—one synchronized with the original carrier and one shifted by 90 degrees. Low-pass filters then remove high-frequency components, yielding the raw coordinate values \((I_{rx}, Q_{rx})\) of the received symbol.

2. Decision Slicing and Distance Calculation

Because of noise, the measured point \((I_{rx}, Q_{rx})\) rarely lands exactly on an ideal coordinate. The receiver determines the most likely transmitted point using Maximum Likelihood Detection. It calculates the Euclidean distance between the received point and all possible valid constellation points:

\[d = \sqrt{(I_{rx} - I_{ideal})^2 + (Q_{rx} - Q_{ideal})^2}\]

The ideal point with the shortest geometric distance to the received coordinate is selected as the intended symbol. In hardware, this is often implemented efficiently using decision boundaries (slicers) that divide the constellation grid into regions.

3. Binary De-Mapping Using Gray Coding

Once the nearest ideal point is identified, the receiver translates the coordinate back into its binary representation using a predefined lookup table or mapping logic.

Standard QAM systems employ Gray coding for this mapping. Under Gray coding, adjacent constellation points differ by only a single binary digit (e.g., 00 is next to 01, not 11). This ensures that if noise causes a symbol to cross a decision boundary into an adjacent region, it results in only a single bit error rather than multiple corrupted bits.

4. Serialization

Finally, the recovered parallel bits from each symbol (e.g., 4 bits from a 16-QAM point) are loaded into a shift register or digital buffer and converted into a continuous serial binary stream (0s and 1s) ready for downstream digital processing.