Rice distribution

In the 2D plane, pick a fixed point at distance ν from the origin. Generate a distribution of 2D points centered around that point, where the x and y coordinates are chosen independently from a Gaussian distribution with standard deviation σ (blue region). If R is the distance from these points to the origin, then R has a Rice distribution.
Probability density function
Rice probability density functions σ = 1.0
Cumulative distribution function
Rice cumulative distribution functions σ = 1.0
Notation Rice(ν,σ)
Parameters ν≥0, distance between the reference point and the center of the bivariate distribution,
σ≥0, scale
Support x∈[0,∞)
PDF xσ2exp⁡(−(x2+ν2)2σ2)I0(xνσ2)
CDF

1−Q1(νσ,xσ)

where Q1 is the Marcum Q-function
Mean σπ/2L1/2(−ν2/2σ2)
Variance 2σ2+ν2−πσ22L1/22(−ν22σ2)
Skewness (complicated)
Excess kurtosis (complicated)

In probability theory, the Rice distribution or Rician distribution (or, less commonly, Ricean distribution) is the probability distribution of the magnitude of a circularly symmetric bivariate normal random variable, possibly with non-zero mean (noncentral). It was named after Stephen O. Rice (1907–1986).

Characterization

The probability density function is

f(x∣ν,σ)=xσ2exp⁡(−(x2+ν2)2σ2)I0(xνσ2)H(x),

where I0(z) is the modified Bessel function of the first kind with order zero, and H(x) is the Heaviside unit step.

In the context of Rician fading, the distribution is often also rewritten using the shape parameter K=ν22σ2, defined as the ratio of the power contributions by line-of-sight path to the remaining multipaths, and the scale parameter Ω=ν2+2σ2, defined as the total power received in all paths.

The characteristic function of the Rice distribution is given as:

χX(t∣ν,σ)=exp⁡(−ν22σ2)[Ψ2(1;1,12;ν22σ2,−12σ2t2)+i2σtΨ2(32;1,32;ν22σ2,−12σ2t2)],

where Ψ2(α;γ,γ′;x,y) is one of Horn's confluent hypergeometric functions with two variables and convergent for all finite values of x and ⁠y⁠. It is given by:

Ψ2(α;γ,γ′;x,y)=∑n=0∞∑m=0∞(α)m+n(γ)m(γ′)nxmynm!n!,

where

(x)n=x(x+1)⋯(x+n−1)=Γ(x+n)Γ(x)

is the rising factorial.

Properties

Moments

The first few raw moments are:

μ1′=σπ/2L1/2(−ν2/2σ2)μ2′=2σ2+ν2μ3′=3σ3π/2L3/2(−ν2/2σ2)μ4′=8σ4+8σ2ν2+ν4μ5′=15σ5π/2L5/2(−ν2/2σ2)μ6′=48σ6+72σ4ν2+18σ2ν4+ν6

and, in general, the raw moments are given by

μk′=σk2k/2Γ(1+k/2)Lk/2(−ν2/2σ2).

Here ⁠Lq(x)⁠ denotes a Laguerre polynomial:

Lq(x)=Lq(0)(x)=M(−q,1,x)=1F1(−q;1;x)

where M(a,b,z)=1F1(a;b;z) is the confluent hypergeometric function of the first kind. When ⁠k⁠ is even, the raw moments become simple polynomials in ⁠σ⁠ and ⁠ν⁠, as in the examples above.

For the case ⁠q=1/2⁠:

L1/2(x)=1F1(−12;1;x)=ex/2[(1−x)I0(−x2)−xI1(−x2)].

The second central moment, the variance, is

μ2=2σ2+ν2−(πσ2/2)L1/22(−ν2/2σ2).

Note that L1/22(⋅) indicates the square of the Laguerre polynomial ⁠L1/2(⋅)⁠, not the generalized Laguerre polynomial ⁠L1/2(2)(⋅)⁠.

Limiting cases

For large values of the argument, the Laguerre polynomial becomes

limx→−∞Lν(x)=|x|νΓ(1+ν).

It is seen that as ⁠ν⁠ becomes large or ⁠σ⁠ becomes small, the mean becomes ⁠ν⁠ and the variance becomes ⁠σ2⁠.

The transition to a Gaussian approximation proceeds as follows. From Bessel function theory we have

Iα(z)→ez2πz(1−4α2−18z+⋯) as z→∞

so, in the large xν/σ2 region, an asymptotic expansion of the Rician distribution:

f(x,ν,σ)=xσ2exp⁡(−(x2+ν2)2σ2)I0(xνσ2) is xσ2exp⁡(−(x2+ν2)2σ2)σ22πxνexp⁡(2xν2σ2)(1+σ28xν+⋯)→1σ2πexp⁡(−(x−ν)22σ2)xν, as xνσ2→∞

Moreover, when the density is concentrated around ν and |x−ν|≪σ because of the Gaussian exponent, we can also write x/ν≈1 and finally get the Normal approximation

f(x,ν,σ)≈1σ2πexp⁡(−(x−ν)22σ2),νσ≫1

The approximation becomes usable for ⁠ν/σ>3⁠.

Parameter estimation (Koay inversion technique)

There are three different methods for estimating the parameters of the Rice distribution, (1) method of moments, (2) method of maximum likelihood, and (3) method of least squares.[citation needed] In the first two methods the interest is in estimating the parameters of the distribution, ⁠ν⁠ and ⁠σ⁠, from a sample of data. This can be done using the method of moments, e.g., the sample mean and the sample standard deviation. The sample mean is an estimate of ⁠μ1'⁠ and the sample standard deviation is an estimate of ⁠μ21/2⁠.

The following is an efficient method, known as the "Koay inversion technique". for solving the estimating equations, based on the sample mean and the sample standard deviation, simultaneously. This inversion technique is also known as the fixed point formula of SNR. Earlier works on the method of moments usually use a root-finding method to solve the problem, which is not efficient.

First, the ratio of the sample mean to the sample standard deviation is defined as ⁠r⁠, i.e., ⁠r=μ1'/μ21/2⁠. The fixed point formula of SNR is expressed as

g(θ)=ξ(θ)[1+r2]−2,

where θ is the ratio of the parameters, i.e., ⁠θ=ν/σ⁠, and ξ(θ) is given by:

ξ(θ)=2+θ2−π8exp⁡(−θ2/2)[(2+θ2)I0(θ2/4)+θ2I1(θ2/4)]2,

where I0 and I1 are modified Bessel functions of the first kind.

Note that ξ(θ) is a scaling factor of σ and is related to μ2 by:

μ2=ξ(θ)σ2.

To find the fixed point, ⁠θ∗⁠, of ⁠g⁠, an initial solution is selected, ⁠θ0⁠, that is greater than the lower bound, which is θlower bound=0 and occurs when r=π/(4−π) (Notice that this is the r=μ1'/μ21/2 of a Rayleigh distribution). This provides a starting point for the iteration, which uses functional composition,[clarification needed] and this continues until |gi(θ0)−θi−1| is less than some small positive value. Here, gi denotes the composition of the same function, ⁠g⁠, i times. In practice, we associate the final θn for some integer n as the fixed point, ⁠θ∗⁠, i.e., ⁠θ∗=g(θ∗)⁠.

Once the fixed point is found, the estimates ν and σ are found through the scaling function, ⁠ξ(θ)⁠, as follows:

σ=μ21/2ξ(θ∗),

and

ν=(μ1' 2+(ξ(θ∗)−2)σ2).

To speed up the iteration even more, one can use the Newton's method of root-finding. This particular approach is highly efficient.

Applications

See also