2
$\begingroup$

Given an arbitrary 3-qubit state $\sum_{xyz} c_{xyz}|xyz\rangle$, is there a circuit (possibly with measurement) that creates the state $\sum_{xy} c_{xyy}|x\rangle$, up to a normalization constant?

As a weaker question, is the above circuit possible, given that the initial state can be decomposed as a tensor product of the first two qubit and the third? This would allow an encoding of a quantum gate with 2 qubits.

I see that measuring the last two qubits in the Bell basis has a chance of succeeding. But for 3 out of 4 cases, although I know I failed, I cannot correct for it as in the quantum teleporting case, because the gates needed for correction is unknown. Is this task impossible? How can I prove it?

$\endgroup$
3
  • $\begingroup$ Typo? Should be $|xy\rangle$ instead of $|x\rangle$? $\endgroup$
    – narip
    Sep 2, 2022 at 12:45
  • $\begingroup$ @narip No, $y$ should be contracted away. $\endgroup$
    – Trebor
    Sep 2, 2022 at 13:42
  • $\begingroup$ What's the difference between $\sum_{xy} c_{xyy}|x\rangle$ and $\sum_{x} \tilde c_{x}|x\rangle$? $\endgroup$
    – narip
    Sep 3, 2022 at 2:39

1 Answer 1

1
$\begingroup$

We know that every quantum operation can be described by a completely positive trace-preserving (CPTP) linear map acting on density matrices. You want a map that sends

$$ \sum_{xyz,\tilde x \tilde y\tilde z} c_{xyz}c_{\tilde x\tilde y\tilde z}^*|xyz\rangle\langle \tilde x\tilde y\tilde z|\rightarrow \sum_{xy,\tilde x \tilde y} c_{xyy}c_{\tilde x\tilde y\tilde y}^*|xyy\rangle\langle \tilde x\tilde y\tilde y| $$

Unfortunately, as I've written it, the map isn't even trace-preserving! We can fix that by adding a normalization constant:

$$ \sum_{xyz,\tilde x \tilde y\tilde z} c_{xyz}c_{\tilde x\tilde y\tilde z}^*|xyz\rangle\langle \tilde x\tilde y\tilde z|\rightarrow \frac{1}{\sum_{xy}|c_{xyy}|^2}\sum_{xy,\tilde x \tilde y} c_{xyy}c_{\tilde x\tilde y\tilde y}^*|xyy\rangle\langle \tilde x\tilde y\tilde y| $$

Now we have a trace-preserving operation. But it's no longer linear! Overall, it's not possible to have this operation be both linear and trace-preserving, so it is not a valid quantum operation.

$\endgroup$

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.