A unitary channel is a channel $\mathcal{U}$ of the following form: $\mathcal{U}(\rho) = U\rho U^{\dagger}$.

A mixed unitary channel is a channel $\mathcal{U}_m$ of the form: $\mathcal{U}_m(\rho) = \sum_{k=1}^n p_kU_k\rho U_k^{\dagger}$, where each $U_k$ is a unitary matrix.

So given a general quantum channel, determining whether or not it is mixed unitary is an NP-hard problem. Is there any easy test to check if it is unitary?

  • 1
    $\begingroup$ Can you specify exactly what you mean by an 'isometric' quantum channel? $\endgroup$
    – JSdJ
    Nov 11, 2021 at 12:52
  • $\begingroup$ I have mentioned it in the edits. Thanks for pointing it out. $\endgroup$ Nov 11, 2021 at 13:20
  • 1
    $\begingroup$ In the title, you ask "how one can check if a given isometric channel is unitary" but then in the question you state that "an isometric channel is a unitary channel". Are you just wondering when a mixed unitary channel has a single term in the sum? $\endgroup$
    – Condo
    Nov 11, 2021 at 14:59
  • $\begingroup$ Thanks for pointing out the confusion. I have edited it and hope it is clear now. Given a quantum channel, I was wondering how we can check if it is unitary with only one term in the sum (not mixed unitary). $\endgroup$ Nov 11, 2021 at 15:11
  • $\begingroup$ which problem exactly are you saying is NP-hard? Checking whether a given channel is isometric surely isn't no? Just check the rank of the Choi, as per the answer. And an isometric channel is a unitary channel iff the input and output dimensions are the same (I'm assuming the $A$ in your post is also an isometry, otherwise the map you get is not a (CPTP) channel) $\endgroup$
    – glS
    Nov 11, 2021 at 15:45

1 Answer 1


Compute the rank of the Choi matrix $C_\Phi$. For any quantum channel $\Phi$, the Choi matrix will be rank 1 if and only if $\Phi$ can be written in the form $U\rho U^\dagger$ for a unitary $U$.

  • $\begingroup$ Thanks, this clarifies my doubt. Could you refer any material so that I can read more on this? $\endgroup$ Nov 11, 2021 at 15:26
  • 3
    $\begingroup$ @HarikrishnanSV Watrous' book, cs.uwaterloo.ca/~watrous/TQI, is always a good reference $\endgroup$
    – glS
    Nov 11, 2021 at 15:41
  • 1
    $\begingroup$ @HarikrishnanSV as pointed to by glS in the comment above, Watrous' book is an excellent reference and should provide all the relevant information and background. I believe chapters 2 and 4 discuss quantum channels and mixed-unitary channels. $\endgroup$
    – Condo
    Nov 11, 2021 at 16:47
  • $\begingroup$ Thanks. I will check it out. $\endgroup$ Nov 11, 2021 at 17:11

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.