Usually in radio systems in order to move the frequency response of the message signal to a new frequency band centered at f, the message signal is mixed with $\cos(2\pi ft)$.

What would change in frequency if we multiply the message signal to $\sin(2\pi ft)$ instead of $\cos(2\pi ft)$?


It does the same thing (i.e. shift the frequency to another band), though mathematically it differs a bit. Fourier transforms of the two choices are as shown below:

$$ \begin{array}{|c|c|}\hline f(x)\cos(ax) & \displaystyle\frac{\hat{f}\left(\xi-\frac{a}{2\pi}\right)+\hat{f}\left(\xi+\frac{a}{2\pi}\right)}{2}\\ \hline f(x)\sin(ax) & \displaystyle\frac{\hat{f}\left(\xi-\frac{a}{2\pi}\right)-\hat{f}\left(\xi+\frac{a}{2\pi}\right)}{2i}\\\hline \end{array} $$

The negative frequencies not only get shifted, but also flipped in polarity

Now for the derivation of these formulas: Using the Great Mr. Euler's identity: $e^{iax} = \cos(ax) + i \sin(ax)$

$$\cos(ax) = \frac{e^{iax} + e^{-iax}}{2}$$ and $$\sin(ax) = \frac{e^{iax} - e^{-iax}}{2i}$$

\begin{align} \textrm{F.T.}\left\{f(x) \cos(ax)\right\} &= \int\limits_{-\infty}^{\infty} f(x) \cos(ax) e^{-i2\pi\xi x} dx\\ &= \int\limits_{-\infty}^{\infty} f(x) \frac{e^{iax} + e^{-iax}}{2} e^{-i2\pi\xi x} dx\\ &= \frac{\int\limits_{-\infty}^{\infty} f(x) e^{-i2\pi(\xi + \frac{a}{2\pi} )x} dx + \int\limits_{-\infty}^{\infty} f(x) e^{-i2\pi(\xi - \frac{a}{2\pi})x} dx}{2}\\ &= \frac{\hat{f}(\xi + \frac{a}{2\pi} ) + \hat{f}(\xi - \frac{a}{2\pi} )}{2} \end{align}

You can derive the formula for $\sin$.

  • 1
    $\begingroup$ Nice; got my upvote! I think this question would benefit even more if you'd add the relation between $e^{jx}$ and $\cos$ and $\sin$ explicitly, and hence explained how these transform table entries happen. $\endgroup$ – Marcus Müller Sep 1 '16 at 21:06
  • $\begingroup$ Ping: Srini, I think you can get a lot more upvotes if you actually derive these relationships in your question! $\endgroup$ – Marcus Müller Sep 2 '16 at 7:23
  • $\begingroup$ @Marcus, I am a bit latex-lazy, so was postponing the edit till the weekend :) Your ping made me cut short my laziness. $\endgroup$ – Srini Sep 2 '16 at 14:46
  • $\begingroup$ I wish I could upvote twice :) $\endgroup$ – Marcus Müller Sep 2 '16 at 18:41

Short answer: a phase shift of 90°, since $\cos(t) = \sin(t+\frac\pi2)$.


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.