Contents

Rayleigh DistributionEdit

The Rayleigh distribution is a continuous probability distribution for nonnegative real numbers that arises naturally as the distribution of the magnitude of a two-dimensional vector whose components are independent and identically distributed normal random variables with zero mean. It is named after Lord Rayleigh, who studied the statistical properties of scattered waves and other phenomena in the late 19th century. In practical terms, the Rayleigh distribution provides a simple, tractable model for a broad class of envelope-like measurements encountered in engineering, physics, and related fields. The distribution is fully described by a single scale parameter, commonly denoted σ, which sets the spread of the values.

In engineering and the physical sciences, the Rayleigh distribution is a workhorse model because many real-world processes yield nonnegative magnitudes whose squared value is proportional to the sum of squares of two independent Gaussian components. This makes the Rayleigh distribution a natural choice for modeling the amplitude of signals subject to two-dimensional Gaussian noise, as well as for modeling wind speed under certain assumptions about the horizontal wind components. The mathematical tractability of Rayleigh models—closed forms for the density, cumulative distribution, and many moments—helps engineers and scientists devise reliable estimators, simulate scenarios, and perform analytic performance assessments.

Formal definition

Let R be a nonnegative random variable. R follows the Rayleigh distribution with scale parameter σ > 0 if its probability density function (PDF) is

f(x; σ) = (x / σ^2) exp(-x^2 / (2 σ^2)), for x ≥ 0,

and f(x; σ) = 0 for x < 0. The CDF is

F(x; σ) = 1 − exp(-x^2 / (2 σ^2)), for x ≥ 0.

Key moments and characteristics include: - Mean: E[R] = σ sqrt(π/2) - Median: med(R) = σ sqrt(2 ln 2) - Mode: x = σ - Variance: Var(R) = (4 − π)/2 · σ^2 - Second moment: E[R^2] = 2 σ^2

A compact interpretation is that if X and Y are independent N(0, σ^2) random variables, then R = sqrt(X^2 + Y^2) has the Rayleigh distribution with scale σ. Equivalently, the squared radius R^2 follows an exponential distribution with mean 2 σ^2. The Rayleigh distribution is a special case of the more general Weibull family (shape parameter k = 2) and is related to the chi distribution with two degrees of freedom.

Relationship to other distributions

  • Link to the two-dimensional Gaussian: as noted above, R arises as the magnitude of a 2D Gaussian vector with zero mean and equal variance.
  • Connection to the chi distribution: R ∼ χ2(σ) in the sense that it is the chi distribution with 2 degrees of freedom, up to the chosen parameterization.
  • Connection to the exponential distribution: R^2/(2 σ^2) ∼ Exp(1), so many results for exponential variables transfer to squared Rayleigh variables.
  • Relation to the Rice distribution: the Rice distribution generalizes Rayleigh by allowing a nonzero mean in the underlying Gaussian components; Rayleigh is the special case with zero mean, corresponding to Rice with a zero direct-path component.

Parameter estimation and inference

Estimating the scale parameter σ from data is a standard problem in statistical inference. For a sample {R1, R2, ..., Rn} drawn from a Rayleigh(σ) distribution: - Maximum likelihood estimator (MLE) of σ^2 is

σ̂^2_MLE = (1 / (2n)) ∑_{i=1}^n Ri^2.

This estimator is unbiased for σ^2. A natural estimator for σ is σ̂ = sqrt(σ̂^2_MLE). Method-of-moments estimation aligns with this result because E[R^2] = 2 σ^2, so equating the sample mean of Ri^2 to 2 σ^2 yields the same σ̂^2.

Goodness-of-fit assessments for a Rayleigh model often involve comparing empirical and theoretical CDFs, Q-Q plots against the Rayleigh quantiles, or likelihood-based tests against competing models (e.g., Rice, Weibull, or Gamma families). In practice, the choice of σ reflects the scale of the underlying two-dimensional Gaussian components, so estimation is typically robust when the data reasonably arise from an envelope of many small, independent contributions.

Applications

  • Signal processing and communications: the envelope of a complex baseband signal under Rayleigh fading—where the in-phase and quadrature components are independent Gaussian noise with zero mean—follows a Rayleigh distribution. This makes Rayleigh modelling central to analyses of wireless link reliability and capacity under multipath propagation.
  • Radar and sonar: strength measurements often have Rayleigh-like envelopes under appropriate scattering assumptions, aiding detection theory and performance prediction.
  • Wind speed modeling: under certain climatological conditions, the speed of wind can be described by a Rayleigh distribution, reflecting the combination of horizontal wind components. This usage is common in meteorology and risk assessment.
  • Mechanical and materials engineering: many nonnegative measurement processes that are the result of summing two independent fluctuating contributions can be well approximated by Rayleigh envelopes.
  • Rice and related models: when a deterministic (nonzero-mean) component is present in the underlying 2D Gaussian fields, the envelope becomes a Rice distribution; Rosetta Stone between two common envelope models in statistical signal processing.

Simulation and generation

Rayleigh random variables can be generated in multiple ways: - Draw two independent N(0, σ^2) components and compute R = sqrt(X^2 + Y^2). - Use the relation R^2 ≡ −2 σ^2 ln(U) with U ∼ Uniform(0, 1), then R = sqrt(−2 σ^2 ln U). These methods are standard in simulations of communication systems, sensor networks, and physical processes requiring envelope modelling.

See also