0
$\begingroup$

I am a beginer in signal processing. I am currently studying a signal $s(t) = \cos(h(t))$, and I have to calculate its autocorellation. Since $h(t)$ is periodic of period $T_s$, I calculated the value of

$$R_s(\tau) = \frac{1}{T_s} \int_0^{T_s} \cos(h(t)) \cos(h(t-\tau)) dt$$

Honestly, this calculation is not easy (since $h(t)$ is also a cosinus), and in order to simplify my final document, I would like to use complex representation by writing $s(t) = Re\left\{ e^{ih(t)} \right\}$. However, This cannot work for the autocorrelation since $$Re\left\{ e^{ih(t)} e^{-ih(t-\tau)} \right\} = s(t)s(t-\tau) + Re\left\{ \sin(h(t)) \sin(h(t-\tau)) \right\}$$

So, do you know a way to calculate $R_s$ by calculating $\frac{1}{T_s} \int_0^{T_s} e^{ih(t)} e^{-ih(t-\tau)} dt$ ?

I was thinking about a relation like $R_s(\tau) = \frac{1}{2} Re \left\{\frac{1}{T_s} \int_0^{T_s} e^{ih(t)} e^{-ih(t-\tau)} dt \right\}$, but I cannot prove it :(

$\endgroup$
1
  • $\begingroup$ Is this a cold remedy? $\endgroup$ Commented Apr 23, 2020 at 16:42

1 Answer 1

2
$\begingroup$

It's way easier to use the complex representation of cosine not involving the real part operator. In your case this translates into expressing

$$ \cos(h(t))=\frac{e^{jh(t)} + e^{-jh(t)}}{2} $$

Note that this yields an autocorrealtion of $$ R_s(\tau)=\int_{0}^{T_s} \frac{e^{jh(t)} + e^{-jh(t)}}{2}\frac{e^{jh(t-\tau)} + e^{-jh(t-\tau)}}{2} dt $$

$\endgroup$

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.