1
$\begingroup$

Given two stochastic processes, $X[n]$ and $Y[n]$, both being WSS (wide state stationary) and independents. What would be the Average and Autocorrelation function of $Z[n] = Y[n] X[n]$?

Is the resulting $Z[n]$ process also WSS? If so, can the PSD (power spectral density) be calculated?

Thank you.

$\endgroup$

1 Answer 1

1
$\begingroup$

Let $(X_n)$ and $(Y_n)$ be independent wide-sense stationary (WSS) processes with averages $\mu_X=EX_n$ and $\mu_Y=EY_n$, covariance functions $C_X(m)=E(X_{n+m}-EX_{n+m})(X_{n}-EX_{n})$ and $C_Y(m)=E(Y_{n+m}-EY_{n+m})(Y_{n}-EY_{n})$, correlation functions $R_X(m)=C_X(m)/C_X(0)$ and $R_Y(m)=C_Y(m)/C_Y(0)$, and power spectral densities (PSDs) $f$ and $g$, so that $R_X(m)=\int_{-\infty}^\infty e^{2\pi i mx} f(x)dx$ and $R_Y(m)=\int_{-\infty}^\infty e^{2\pi i mx} g(x)dx$. Let $Z_n=X_n Y_n$.

Then the process $(Z_n)$ is also WSS, with the average $$\mu_Z=EX_n Y_n=EX_n\,EY_n=\mu_X \mu_Y,$$ covariance function \begin{align*} C_Z(m)&=E(X_{n+m}Y_{n+m}-\mu_X\mu_Y)(X_{n}Y_{n}-\mu_X\mu_Y) \\ &=EX_{n+m}X_{n}\,EY_{n+m}Y_n-\mu_X^2\mu_Y^2\\ &=(C_X(m)+\mu_X^2)(C_Y(m)+\mu_Y^2)-\mu_X^2\mu_Y^2\\ &=C_X(m)C_Y(m)+\mu_Y^2C_X(m)+\mu_X^2C_Y(m), \end{align*} correlation function \begin{equation*} R_Z(m)=\frac{C_Z(m)}{C_Z(0)} =a\,R_X(m)R_Y(m) +b\,R_X(m) +c\,R_Y(m) \end{equation*} with $a:=\frac{C_X(0)C_Y(0)}{C_Z(0)}$, $b:=\frac{C_X(0)\mu_Y^2}{C_Z(0)}$, $c:=\frac{C_Y(0)\mu_X^2}{C_Z(0)}$, and PSD \begin{equation*} h=a f*g+bf+cg, \end{equation*} where $f*g$ is the convolution of $f$ and $g$.

$\endgroup$
2
  • $\begingroup$ Hi Iosif. Thank you so much for your answer. One question: What is the 'm' in the Covariance and Correlation functions meaning? Thanks again. $\endgroup$
    – Sergio
    Mar 19, 2018 at 18:09
  • $\begingroup$ @Sergio : If in your question $n$ stands for any integer, then in my answer $m$ also denotes any integer. If in your question $n$ stands for any nonnegative integer (say), then in my answer $m$ also denotes any nonnegative integer. $\endgroup$ Mar 19, 2018 at 22:07

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

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