A quantum map on a $d$-dimensional space has the general representation: $$ \mathcal{S}(\rho)=\sum_{\alpha,\beta}^{d^2}\chi_{\alpha\beta}\Gamma_{\alpha}\rho \Gamma_{\beta}^{\dagger}, $$ where $\chi$ is the $d^2\times d^2$ process matrix, which is positive semidefinite and trace preserving. The $\{\Gamma_{\alpha}\}$ is an orthonormal matrix basis set, and $tr(\Gamma_{\beta}^{\dagger}\Gamma_{\alpha})=\delta_{\alpha\beta}$.

My question: Why there are $d^2$ orthonormal basis? Are there any examples of two-qubit systems?


2 Answers 2


States lie in Hilbert space $\mathcal{H_S}$.

  • $|\psi\rangle \in \mathcal{H_S}\,.$
  • If you have $n$ qubits, then $\text{dim}(\mathcal{H_S}) = 2^n = d\,.$

Operators, density operators lie in the bounded operator space of $\mathcal{H}_S$.

  • $\rho \in \mathcal{B}(\mathcal{H}_S)\,. $
  • $\text{dim}(\mathcal{B}(\mathcal{H}_S)) = d \times d = d^2\,.$

Kraus operators are operators acting on $\mathcal{H_S}$. You have $d^2$ degrees of freedom, which requires $d^2$ basis to describe it.

If your map $\mathcal{E}$ is $$ \mathcal{E}(\rho) = \sum_{n} K_{n} \rho K_{n}^{\dagger}\,, $$

Then you can write your Kraus operators in $\{\Gamma_{\alpha}\}$ basis as $$ K_{n} = \sum_{\alpha=0}^{d^2-1} b_{n\alpha} \Gamma_{\alpha}\,, $$

where $b_{n\alpha}$ are the coefficients for the linear combination using the new basis; therefore, you will have total $d^2$ terms in the summation. $K_n$ is a $d \times d $ matrix.

Now, rewriting our map in terms of Kraus operators in the $\{\Gamma_{\alpha}\}$ basis gives us

$$ \begin{align} \mathcal{E}(\rho) &= \sum_{n} K_{n} \rho K_{n}^{\dagger}\\ &= \sum_n \bigg(\sum_{\alpha=0}^{d^2-1} b_{n\alpha} \Gamma_{\alpha} \bigg) \rho \bigg(\sum_{\beta=0}^{d^2-1} b_{n\beta}^* \Gamma_{\beta}^{\dagger}\bigg)\\ &= \sum_{\alpha=0,\beta=0}^{d^2-1, d^2-1} \big(\sum_n b_{n\alpha} \cdot b_{n\beta}^* \big) \Gamma_{\alpha} \rho \Gamma_{\beta}^{\dagger}\\ &= \sum_{\alpha=0,\beta=0}^{d^2-1, d^2-1} \chi_{\alpha \beta}\Gamma_{\alpha} \rho \Gamma_{\beta}^{\dagger} \\ \end{align} $$

So, as you can see above, Kraus operators being $d \times d$ results in $\chi$ having $d^2 \times d^2$ degrees of freedom. The process matrix is just

$$ \chi = bb^{\dagger} $$

where $b$ is the matrix with linear combination coefficients for your Kraus operators in $\{\Gamma_{\alpha}\}$ basis.


A slight rewording of what's already been said in the other answer: if the underlying space $\mathcal H$ has dimension $d$, then the (real) vector space of Hermitian operators acting on $\mathcal H$ has dimension $d^2$. The elements of the $\chi$ matrix are the coefficients attached to the terms $\Gamma_\alpha \rho \Gamma_\beta^\dagger$ for all $\alpha,\beta$, and $\Gamma_\alpha$ are a basis of Hermitian operators for the space. Thus these indices go from $1$ to $d^2$, and thus there's $d^2\times d^2$ pairs of such indices.


Your Answer

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

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