I was trying to solve this question : enter image description here

I respresented $x(t) = u(t+1)-u(t-1)$

writing the convolution as $[u(t+1)-u(t-1)]*h(t) = y(t)$

I then used the property of differentiation to convert from the step to the impulse function :

$[\delta(t+1) -\delta(t-1)]*h(t) = y'(t)$

$\implies h(t+1) - h(t-1) = y'(t)$

However , I am now stuck here. Moreover , the fact that the width of $y(t)$ is the same as that of $x(t)$ , bugs me a little.

How can I find the impulse response $h(t)$ using the given info ?

  • $\begingroup$ Could you also show the inputs $x_2(t)$ and $x_3(t)$? $\endgroup$
    – Matt L.
    Feb 28, 2021 at 17:29
  • $\begingroup$ @MattL. made the necessary edits $\endgroup$
    – Starboy
    Feb 28, 2021 at 17:31
  • $\begingroup$ Are you familiar with Fourier transforms? $\endgroup$
    – cjferes
    Mar 1, 2021 at 6:46
  • $\begingroup$ @cjferes: I'm afraid that this won't help much in this case. $\endgroup$
    – Matt L.
    Mar 1, 2021 at 10:11

1 Answer 1


This exercise is meant to help the student appreciate the fact that if the response $y_1(t)$ of an LTI system to an input $x_1(t)$ is known, then the response to an input


is given by


which is a direct consequence of linearity and time-invariance.

Consequently, if an input signal can be expressed in the form given by $(1)$, it is not necessary to compute the system's impulse response in order to determine the output signal.

Also note that this exercise is almost a copy of problem P3.8 in this MIT Signals and Systems problem set. Only the input signals have been changed.

It now appears that the given signal $x_2(t)$ can't be expressed in the form $(1)$, whereas signal $x_3(t)$ can. I believe that this a mistake in the exercise, because as far as I can see, any other method to determine the solution is beyond what can be expected from a student at this level.

  • $\begingroup$ I am just not able to represent $x_2(t)$ as an linear combination through just time shifting and amplitude scaling . Seems as if time scaling would be required but that would change the response to $x_1(t)$ $\endgroup$
    – Starboy
    Feb 28, 2021 at 17:56
  • 1
    $\begingroup$ @Starboy: If that is the case then it's a mistake in the exercise. Check problem P3.8 of these MIT signals and systems problems. It's too close to be a coincidence. $\endgroup$
    – Matt L.
    Feb 28, 2021 at 19:03

Your Answer

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

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