Phase modulation (PM) is a form of modulation in which an information-bearing signal controls the phase of a carrier wave relative to a reference. In ideal analog PM, the phase deviation is proportional to the instantaneous value of the message, while the carrier’s amplitude remains constant. Together with frequency modulation (FM), it belongs to the class of angle modulation. Phase modulation is used in telecommunications, optical systems, measurement, and electronic sound synthesis; its digital counterpart, phase-shift keying (PSK), represents symbols through selected carrier phases.
Mathematical description
An ideal phase-modulated sinusoidal carrier can be written as
where:
- is the constant carrier amplitude;
- is the unmodulated carrier frequency;
- is a reference phase;
- is the message signal;
- is the phase sensitivity, measured in radians per unit of the message.
The information-bearing phase deviation is therefore
The phase in this expression is an unwrapped phase: it can increase or decrease beyond a single revolution even though a sinusoidal waveform repeats every radians. That periodicity has practical consequences for phase measurement and receiver ambiguity.
The instantaneous frequency is the time derivative of the total phase, divided by :
Thus, phase modulation generally also produces frequency variation. A constant message value produces a fixed phase offset, whereas a changing message produces a temporary frequency deviation. The distinction between PM and FM concerns which property is directly proportional to the message, not whether the other property changes.
Relationship to frequency modulation
In ideal frequency modulation,
where is the frequency sensitivity in hertz per unit of the message. The frequency deviation is directly proportional to , while the phase deviation is proportional to its integral.
This relationship permits one modulation method to be generated using the other:
- Integrating a message before applying it to a phase modulator produces FM, with suitable scaling.
- Differentiating a message before applying it to a frequency modulator produces PM, again with suitable scaling.
For a sinusoidal message,
the peak phase deviation, or PM modulation index, is
The peak frequency deviation is
At fixed message amplitude, ideal PM therefore has a phase deviation independent of message frequency, but a frequency deviation that increases with message frequency. In FM, fixed message amplitude instead gives a fixed frequency deviation, with phase deviation decreasing as message frequency increases.
A single-tone PM waveform can have the same spectrum as a single-tone FM waveform with the same modulation index, apart from phase relationships. Their differing responses become especially important for messages containing multiple frequencies.
Spectrum and bandwidth
For sinusoidal phase modulation, the waveform may be expressed as
with the constant reference phase omitted. Its spectrum contains a component at and sidebands at
The sideband amplitudes are governed by Bessel functions of the first kind, . Unlike conventional sinusoidal amplitude modulation, ideal sinusoidal PM generally produces infinitely many sidebands, although sufficiently high-order components become very small.
When , the narrowband approximation is
Only the carrier and first pair of sidebands are substantial in this approximation. As the modulation index increases, higher-order sidebands become important. The spectral component at the original carrier frequency can even vanish at particular modulation indices, without making the total transmitted power vanish: energy has been redistributed among the sidebands.
A common engineering estimate of occupied bandwidth is Carson’s rule:
where is the peak instantaneous frequency deviation and is the highest significant message frequency. For a sinusoidal PM signal, this becomes
This is an approximate bandwidth estimate, not a strict spectral boundary. For a general PM message, frequency deviation depends on the maximum rate of change of the message:
Consequently, message amplitude alone does not determine PM bandwidth. Rapid transitions and high-frequency content can substantially broaden the spectrum.
Generation
Electronic phase modulators
A phase modulator changes the phase of an oscillator-derived signal in response to a control voltage or digital command. Practical approaches include voltage-controlled phase-shifting networks and vector modulators.
An in-phase and quadrature implementation uses two carriers separated by a quarter cycle. For a desired phase deviation ,
where
Ideally, , giving a constant-amplitude output. This representation is widely used in digital signal processing and programmable radio systems.
A small-angle quadrature modulator can also approximate PM by adding a message-dependent quadrature component to a strong carrier. Without further correction, this method produces some amplitude variation and is accurate only over a limited phase range.
Optical phase modulators
In optics, an applied electric field can change a material’s refractive index and thereby change the phase accumulated by light passing through it. Electro-optic modulators are used with lasers in optical communication and laboratory measurement.
For a uniform interaction region, a refractive-index change over length produces approximately
where is the vacuum wavelength. Relevant device properties include modulation bandwidth, drive voltage, insertion loss, and unwanted amplitude modulation.
Detection and phase recovery
Recovering a PM message requires measuring the received phase relative to an appropriate reference.
A coherent receiver mixes the received signal with in-phase and quadrature local references. The resulting complex baseband signal ideally has the form
where includes oscillator mismatch and channel-induced phase changes. An angle estimator can recover the phase modulo . When the message requires a continuous phase trajectory, phase unwrapping is needed to reconstruct it.
A phase-locked loop can track carrier phase. Its response is shaped by the loop bandwidth, message dynamics, and noise. The appropriate recovered output depends on the receiver architecture: a phase estimate represents phase modulation, while a frequency-related control signal may require integration to recover the original PM message.
An alternative is to pass the signal through a frequency discriminator and integrate its output. Since a discriminator responds to instantaneous frequency, its output for PM is proportional to the message derivative before integration.
Practical receivers must handle carrier-frequency error, phase ambiguity, oscillator drift, and cycle slips. Strong noise or an excessively rapid phase trajectory can cause a tracking system to lose synchronization.
Digital phase modulation
In PSK, digital symbols select among a finite set of carrier phases. Common forms include:
- Binary phase-shift keying (BPSK): two phases separated by radians.
- Quadrature phase-shift keying (QPSK): four phases, allowing two bits per symbol.
- Higher-order PSK: larger phase alphabets, increasing bits per symbol while reducing angular separation between neighboring symbols.
- Differential PSK: information is represented by changes between successive symbol phases rather than by their absolute phases alone.
Digital PSK belongs to the broader family of phase modulation, but its symbol mapping and detection are usually treated separately from analog PM’s proportional message-to-phase relationship.
An ideal PSK model may have constant amplitude. Actual transmitted signals, however, depend on pulse shaping and the path followed between symbol states. Filtering and some transition schemes produce envelope variation. Constant-envelope behavior should therefore not be assumed for every implementation labeled PSK.
In continuous-phase modulation, phase remains continuous across symbol boundaries. Its spectrum and memory depend on the modulation index and frequency-pulse shape. This family is related to phase modulation but is not simply an analog phase modulator driven by abrupt symbol steps.
Applications and development
Phase modulation developed within the broader study of angle modulation in early radio engineering. Phase-modulating stages became important in indirect methods of generating FM, while later digital processing and integrated modulators enabled more flexible control of carrier phase.
Major applications include:
- Digital radio and satellite links. PSK formats carry data using discrete phase states. Signal design balances bandwidth, error probability, synchronization requirements, and amplifier behavior.
- Optical communication. Phase-encoded signals transmitted through optical fiber can be recovered by coherent detection. Phase modulation is also combined with amplitude modulation in more general optical modulation formats.
- Radar. Phase-coded waveforms support pulse compression, allowing a relatively long transmitted pulse to yield fine delay resolution after matched processing. Such techniques are used in radar systems.
- Precision measurement. Deliberately imposed phase modulation creates optical or electrical sidebands used in spectroscopy, interferometric measurements, and frequency-stabilization systems.
- Sound synthesis. In electronic music, phase modulation creates complex harmonic or inharmonic spectra. Some synthesizer methods commonly described as FM are implemented computationally through modulation of an oscillator’s phase.
Limitations and practical distinctions
Ideal PM has a constant envelope, allowing the use of efficient nonlinear electronic amplifiers without the envelope distortion that affects amplitude-varying signals. This advantage depends on the transmitted waveform actually retaining that property. Filtering, hardware imperfections, and amplitude–phase conversion can introduce distortions.
Phase modulation is not immune to noise. Performance depends on received signal-to-noise ratio, modulation parameters, receiver design, and synchronization. Phase noise from oscillators is particularly relevant because it perturbs the same signal property used to carry information.
The mathematical message-to-phase relation is linear, but the transmitted waveform is a nonlinear function of the message. Superposing two messages therefore does not generally produce the sum of their separately phase-modulated waveforms.
Finally, a static phase shift is not by itself an information-bearing modulation process. Phase adjustments in antennas or optical paths may serve beam steering, interference control, or calibration. They constitute communication modulation only when the phase variation is used to encode a message.