I am wondering about an upper bound of the trace function $\operatorname{tr}[ (\mathcal{H}[L]\rho )^2] \leq \delta$ (we assume that $\rho$ is the $N\times N$ density matrix representing the quantum state, $L$ is the $N\times N$ Hermitian operator $L=L^\dagger$, and $\mathcal{H}[L]\rho=L\rho+\rho L-2 \operatorname{tr}(L\rho)\rho$ ). I would like to have some advice from you.

  • 1
    $\begingroup$ What do you want from the upper bound? You could apply Cauchy-Schwarz or Hölder inequalities but without more context I don't know why they'd be more useful to you than the original expression. $\endgroup$
    – Rammus
    Commented May 8, 2023 at 9:08
  • $\begingroup$ I want to make the upper bound $\delta$ does not depend on $\rho$ and tighter. $\endgroup$
    – Kochan
    Commented May 8, 2023 at 17:52
  • $\begingroup$ For completeness sake—as giving an answer is currently impossible because it is not explained which $\delta$ has been found already—using standard techniques one gets the brute force/sub-optimal upper bound $\operatorname{tr}[ (\mathcal{H}[L]\rho )^2]\leq 16\|L\|_\infty^2$ (with $\|L\|_\infty$ the usual operator norm) which holds for all density matrices $\rho$ $\endgroup$ Commented May 27 at 16:38


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.