Suppose we have the Hilbert space $\mathcal{H}_{0}$ describing the states of the system $s$, and the Hilbert space $\mathcal{H}_{e}$ describing the states of the environment. I have seen that [1][2], if the system is initially in the product state $|\psi_{0}\rangle|\chi_{e}\rangle$ we can write the time evolution as $$\hat{U}|\psi_{0}\rangle|\chi_{\mathrm{e}}\rangle=\sum_{k}\left(\hat{E}_{k}|\psi_{0}\rangle\right)\otimes|\chi_{k}\rangle$$ where $$\sum_{k}\hat{E}^{\dagger}_{k}\hat{E}_{k}=\mathbb{1}$$ Taking the partial trace over the environment, we see that the density operator of state $s$ evolves as $$|\psi_{0}\rangle\langle\psi_{0}|=\sum_{k}\hat{E}_{k}|\psi_{0}\rangle\langle\psi_{0}|\hat{E}^{\dagger}_{k}$$ So, from unitary evolution of a system, the Kraus representation of quantum channels on the subsystem can be derived. According to Neumark's theorem, any POVM measurement in a subsystem can be achieved by a POVM measurement in a larger system, a result which is (to my understanding) analogous to the relationship between channels in a subsystem and unitary operations in a larger system. My question is then, is it possible to similarly derive the form of POVMs by considering a projective measurement in the larger system?

I am asking this, because I currently don't fully understand the point of POVMs, and I feel like seeing this relationship may help to enlighten me a bit. Thanks for any help you may be able to provide.

  • $\begingroup$ I don't understand the question. Yes, Neumark 's theorem tells you that any POVM can be represented as a projective measurement in an enlarged space. Is that what you are asking? $\endgroup$
    – glS
    Apr 21, 2023 at 19:36
  • $\begingroup$ What I’m asking is, if we start with a general projective measurement on a space can we derive the form of resulting POVMs on a sub space ? $\endgroup$ Apr 23, 2023 at 15:59
  • 1
    $\begingroup$ depends on what you mean. What is the "resulting POVM on a sub space"? What "subspace" are you referring to? Simply ignoring some outcomes doesn't result in a POVM, if that's what you're asking $\endgroup$
    – glS
    Apr 23, 2023 at 20:10


Your Answer

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

Browse other questions tagged or ask your own question.