Given the following function of random variables

$$f = \left|\sum_{k=1}^{n}{|h_k||g_k|\exp\left( j \theta_k \right)} \right|^2,$$ where $h_1, \cdots, h_n$ and $g_1, \cdots, h_n$ are i.i.d. random variables following the complex Gaussian distribution $\mathcal{CN}(0, \sigma_{h}^2)$ and $\mathcal{CN}(0, \sigma_{g}^2)$ and $\theta_1, \cdots, \theta_n$ $\in [\frac{-\pi}{Q}, \frac{\pi}{Q}]$ are i.i.d. uniformly distributed random variables with probability density function (PDF) given by $\frac{Q}{2\pi}$, where $Q$ is a integer number greater than 0. Additionally, we assume that $h_k$, $g_k$ and $\theta_k$ are independent for all values of $k$.

UPDATED on 10/01/2020

Based on my simulations I know that the PDF of $r$, which is defined as $$ r = \left|\sum_{k=1}^{n}{|h_k||g_k|\exp\left( j \theta_k \right)} \right|, $$ can be accurately approximated by a Gamma random variable even for small values of $n$, and $Q$, $e.g.$, $n=1$ and $Q=4$. Therefore, the PDF of $f = r^2$ can be approximated as $$ P_{n}(f) = \frac{1}{2 \Gamma(\kappa)\Theta^{\kappa}} f^{\left(\frac{\kappa-2}{2} \right)} \exp\left(-\frac{\sqrt{f}}{\Theta}\right), f > 0.$$

However, I'm not being able to find the $\kappa$ and $\Theta$, $i.e.$, shape and scale, parameters of the Gamma random variable.

This problem arises from the study on wireless communications channels and is of great importance to the research community.

  • 1
    $\begingroup$ in one of your earlier questions you mentioned drawing from a Nagakami distribution, rather than from a Gaussian distribution; is there a reason you switched? (I am asking, because that earlier question gave the same motivation from "wireless communication".) $\endgroup$ Commented Jan 5, 2020 at 21:53
  • $\begingroup$ @CarloBeenakker, thanks for your reply. Nakagami channels are only one of several possible channel models. I thought it would be easier to start from the simpler Rayleigh channel, which is described by the absolute values of $g$ and $h$. $\endgroup$ Commented Jan 6, 2020 at 1:09
  • 1
    $\begingroup$ Just to be sure for the definition of the law of $\theta_i$, Is it that $\theta_i\in [0,\frac{2\pi}{Q}]$? $\endgroup$
    – RaphaelB4
    Commented Jan 6, 2020 at 11:13
  • $\begingroup$ @RaphaelB4, $\theta_{k} \in [\frac{-\pi}{Q}, \frac{\pi}{Q}]$. I've just updated the quaestion with this information.Thanks! $\endgroup$ Commented Jan 6, 2020 at 12:28

2 Answers 2


n $\mathbf{\gg}$ 1:

The variable $f=r^2$ is the square of the distance $r$ from the origin after $n$ steps of a random walk on the plane with random direction [$\theta$ uniformly in $(0,2\pi)$] and mean squared step size $s^2$ given by $$s^2=\mathbb{E}(|h|^2||g|^2)=\mathbb{E}(|h|^2|)\mathbb{E}(|g|^2)=\sigma_h^2\sigma_g^2.$$ For $n\gg 1$ the distribution of $r$ is a Maxwell distribution and hence the distribution of $f$ is exponential, $$P_n(r)=\frac{2r}{ns^2}e^{-\frac{r^2}{ns^2}}\Rightarrow P_n(f)=\frac{1}{ns^2}e^{-\frac{f}{ns^2}}.$$ So for $n\gg 1$ the distribution $P_n(f)$ is a Gamma distribution with shape $k=1$ and scale $\theta=n\sigma_h^2\sigma_g^2$.

n $=$ 1:

For $n=1$ one can calculate directly $$P_1(r)=\int_0^\infty d|g|\int_0^\infty d|h|\, \delta(r-|gh|)P(|h|)P(|g|)$$ $$=(\sigma_h\sigma_g)^{-2}\int_0^\infty dh\, \frac{4r}{h} e^{-h^2/\sigma_h^2} e^{-(r/h)^2/\sigma_g^2}=\frac{4r}{\sigma_g^2\sigma_h^2}K_0\left(\frac{2r}{\sigma_g\sigma_h}\right),$$ with $K_0$ a Bessel function.
(I have used that the absolute value $|h|$ when $h$ is $\mathcal{CN}(0, 1)$ has distribution $p(|h|)=2|h|e^{-|h|^2}$.)
So for $n=1$ the distribution of $f=r^2$ is $$P_1(f)=\frac{2}{\sigma_g^2\sigma_h^2}K_0\left(\frac{2\sqrt{f}}{\sigma_g\sigma_h}\right).$$

arbitrary n:

A closed form for $P_n(r)$ exists for arbitrary $n$, see The non-isotropic two-dimensional random walk, by B.C. Barber (1993):

$$P_n(f)=(\sigma_g\sigma_h)^{-n-1}\frac{2f^{(n-1)/2}}{\Gamma(n)}K_{n-1}\left(\frac{2\sqrt f}{\sigma_g\sigma_h}\right).$$

The large-$f$ decay is a stretched exponential $\propto e^{-2\sqrt{f}}$, so this cannot be precisely modeled by a Gamma distribution (which decays exponentially).

anisotropic random walk:

All of this was for an isotropic random walk, with $\theta_k$ at each step uniformly distributed in $(-\pi,\pi)$. The parameter $Q>1$ restricts the angle to $(-\pi/Q,\pi/Q)$, so the scattering is peaked in the forward direction. For $n\gg 1$ the anisotropy can be accounted for by a rescaling of the mean squared step size, $$s_Q^2=\frac{\sigma_h^2\sigma_g^2}{1-\langle \cos\theta\rangle},$$ where $\langle\cdots\rangle$ denotes the angular average. We have $$\langle\cos\theta\rangle=\frac{Q}{2\pi}\int_{-\pi/Q}^{\pi/Q}\cos\theta\,d\theta=\frac{Q}{\pi}\sin(\pi/Q),$$ so we arrive at the distribution $$P_n(f)=\frac{1}{ns_Q^2}e^{-\frac{f}{ns_Q^2}},\;\;s_Q^2=\frac{\sigma_h^2\sigma_g^2}{1-(Q/\pi)\sin(\pi/Q)}.$$ This is a Gamma distribution with shape $k=1$ and scale $\theta=ns_Q^2$.

For $n=1$ the distribution $P_1(f)$ is $Q$-independent, so the result $\propto K_0$ above still applies. For $n>1$ I presume the decay will still be a stretched exponential, but I do not have an exact result as for $Q=1$.

  • 1
    $\begingroup$ for $n\gg 1$ the parameters of the Gamma distribution are shape $=1$ and scale $=n\sigma_h^2\sigma_g^2$. For $n=1$ I find a Bessel function distribution, which for small $f$ diverges logarithmically as $-\ln f$ and for large $f$ decays as a stretched exponential $e^{-\sqrt f}$. So this is definitely not a Gamma distribution, but if you insist you might take the shape $k=1/2$, which gives a reasonable fit (see plot). Interpolating between $k=1/2$ for $n=1$ and $k=1$ for $n\geq 10$, with $\theta=n\sigma_h^2\sigma_g^2$, should then cover the entire range of $n$. Is that satisfactory? $\endgroup$ Commented Jan 6, 2020 at 11:17
  • 1
    $\begingroup$ I've added the $Q>1$ case for $n\gg 1$, which is a Gamma distribution with shape $=1$ and scale $=n\sigma_h^2\sigma_g^2[1-(Q/\pi)\sin(\pi/Q)]^{-1}$. I have also added the exact result for all $n$ for $Q=1$ (in terms of a Bessel function $K_{n-1}$). $\endgroup$ Commented Jan 6, 2020 at 18:49
  • 1
    $\begingroup$ there is not really much more to say: in the limit $n\gg 1$ of many steps the random walk satisfies a diffusion equation, which contains only a single parameter, the diffusion constant $D$, or equivalently, the mean free path $l$. For isotropic scattering in $d$ dimensions $D$ is related to the mean-square-displacement $s^2$ by $s^2=2dD\tau$ (with $\tau$ the time between steps). For anisotropic scattering you should replace $D$ by $D/(1-\langle\cos\theta\rangle)$, or equivalently, replace $l$ by the transport mean free path. $\endgroup$ Commented Jan 7, 2020 at 12:15
  • 1
    $\begingroup$ certainly, once you are in the diffusion regime $n\gg 1$ all you need to know is the mean-square displacement, it does not matter how the step size is distributed. $\endgroup$ Commented Jan 7, 2020 at 14:07
  • 1
    $\begingroup$ if you have more and more forward scattering, for larger and larger $Q$, you will need larger and larger $n$ to reach the diffusion regime, simply because you need backscattering for diffusion and backscattering is suppressed for large $Q$; if $Q$ is large and $n$ is small a better approximation would be to ignore backscattering altogether, and just fix the angles $\theta_k$ at zero. $\endgroup$ Commented Jan 9, 2020 at 15:28

I understand the answer I gave, which involves either exact or asymptotic results, is not quite what the OP was looking for, so let me try to suggest a different approach. The question is: please find parameters $k$ and $\theta$ such that $$p_{k,\theta}(r) = \frac{r^{k-1}}{ \Gamma(k)\theta^{k}} e^{-r/\theta}$$ approximates the distribution of the distance from the origin after $n$ steps in a certain anisotropic random walk on the two-dimensional plane. The random walk has step size distribution $$F(r)=\frac{4r}{\sigma_g^2\sigma_h^2}K_0\left(\frac{2r}{\sigma_g\sigma_h}\right)$$ and scattering angle uniformly distributed in $(-\pi/Q,\pi/Q)$.

The first step would be to eliminate the parameters $\sigma_g$ and $\sigma_h$ by rescaling $$r=R\sigma_g\sigma_h$$ The distribution $P_{n,Q}(R)$ of $R$ no longer depends on $\sigma_g$ or $\sigma_h$. It only contains the two parameters $n$ and $Q$, both $\in\mathbb{N}$, and we wish to fit this with the two-parameter distribution $p_{k,\theta}(R)$, with $k,\theta\in\mathbb{R}$.

The next thing to note is that in the context in which the problem appears, only a few integer values of $Q$ are of relevance. We also have exact asymptotic results for $n\gg 1$, so if we would take 100 different combinations of $Q$ and $n$ we would have covered much of the relevant parameter space. The fitting procedure could be simplified by fitting first for large $R$, to extract $\theta$, and then for small $R$, to extract $k$. I guess this could be automated to a large extent, or perhaps even done by hand.

As an example, I tried this for $Q=1$, when I know the exact $P_{n,1}(r)$. The Mathematica command

Manipulate[ LogPlot[{(4*r^n/Gamma[n])*BesselK[n-1, 2*r], (r^(k-1)/(Gamma[k]*t^k))*Exp[-r/t]},{r,0,20}], {n,2,10,1}, {k,0.1,5}, {t,0.1,5}]

provides for a convenient way to do the fit by sliding the parameters $n$, $k$, and $t\equiv\theta$. Here is a representative output for $n=3,5,8$:

I followed up on the suggestion by the OP to carry out the fit by adjusting the parameters $k$ and $\theta$ such that the first two moments of $R$ agree. For $Q=1$ we have the exact result $$\mathbb{E}(R)=\frac{\sqrt{\pi } \,\Gamma \left(n+\frac{1}{2}\right)}{2 \Gamma (n)},\;\;\mathbb{E}(R^2)=n,$$ to be compared with the Gamma-distribution result $\mathbb{E}(R)=k\theta$, $\mathbb{E}(R^2)=k(1+k)\theta^2$.
I show the comparison for $n=5$, when $k=2.955$ and $\theta=0.654$, see the plot below. It seems this fitting procedure is worse than the unconstrained fit (with $k=3.6$ and $\theta=0.55$) carried out above.

  • $\begingroup$ Dear Carlo, thanks once more for the new answer. What I'm trying to do right now is to find $\kappa$ and $\Theta$ based on the Gamma's distribution definition of $\mu$ and $\sigma^2$, $i.e.$, $\mu = \kappa \Theta$ and $\sigma^2 = \kappa \Theta^2$. The only problem I've found is how to find the expectation of $r$. If you have any jint on that I would be very glad! $\endgroup$ Commented Jan 13, 2020 at 19:20
  • $\begingroup$ If we find two different moments for $Q \neq 1$, for example the mean and variance, we will be able to find $\kappa$ and $\Theta$. I will post an expansion for $r^2$. $\endgroup$ Commented Jan 13, 2020 at 21:42
  • 1
    $\begingroup$ By using the following identity $r^2 = | \sum_{i=1}^{N}{z_{i} e^{\theta_{1}}} |^2 = \sum_{i=1}^{N}{z_{i}^{2}} + 2\sum_{i=1}^{N}{\sum_{j=i+1}^{N}{z_{i}z_{j}\cos(\theta_{i}-\theta_{j})}}$ we can find the expectation of $r^2$. Then, if we find a different expectation we can use the two of them to figure out $\kappa$ and $\Theta$. $\endgroup$ Commented Jan 13, 2020 at 23:59
  • 1
    $\begingroup$ I tried this for $Q=1$. The result is not very encouraging, a global fit produces much better results than the fit of the first two moments. $\endgroup$ Commented Jan 14, 2020 at 11:07
  • 1
    $\begingroup$ for $Q=1$ I have an exact expression for $P(R)$ in terms of the $K_{n-1}$ Bessel function (see my other answer); from that I can immediately calculate all moments of $R$. $\endgroup$ Commented Jan 14, 2020 at 14:38

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.