I have $x(n)$ = {$4,-1,-3,1$} and $h(n)$ = {$2,1,3,5$}.

I would like to know how can I check whether

$h(n)$ is stable filter or not and $h(n)$ is causal filter or not?

  • 1
    $\begingroup$ hey! Your notation isn't quite right, and that makes it impossible for you to answer this yourself :) $h(n)$ is a single value, the value of the signal (or impulse response) $h$ at time $n$. Now, $\{2,1,3,5\}$ is a set of values, and that's not what you wanted. So, first step, figure out for which $n$ the value of $h$ becomes $2$, for which it becomes $1$ and so on. Then check the definition of "causal"! $\endgroup$ Apr 19, 2020 at 9:12
  • $\begingroup$ let say h(n = 0) = 2, h(n = 1) = 1, h(n = 2) = 3, h(n = 3) = 5. $\endgroup$ Apr 19, 2020 at 9:17
  • 2
    $\begingroup$ if you say so, congratulations, you've just answered your own causality question. $\endgroup$ Apr 19, 2020 at 9:23

1 Answer 1


Firstly, you do not need to specify an input $x(n)$ to know whether the system is stable or causal, these are properties of the system itslef, I will interpret $h(n)$ you mention as a array describing the system response.

For causality in time domain perspectivefor an LTI system check if $h(n) =0 $ for all $n < 0$, that is the $h(n)$ is essentially a right sided sequence.

For causality in Z domain perspective equivalently, the Region of Convergence is the right side of the outermost pole (outermost in terms of the highest magnitude pole in z domain)

For stability in time domain perspective, the system response $h(n)$ should be absolutely summable, since you have only finite coefficients, this system is stable, any FIR filter is always stable.

For stability from Z domain perspective, the region of Convergence must contain the unit circle.

For stable and causal system all these conditions hold at the same time.


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