Lean, informalized

Nielsen & Chuang · Chapter 8, Quantum noise and quantum operations · p. 385

Exercise 8.27 - The phase flip operation elements as unitary combinations of the phase damping elements

A unitary that turns phase damping into a phase flip.

Hover over a blue or green term in the text, or a node in the graph, to find it in the other. Select it to read its declaration both in Lean and in words.

The Exercise

Phase damping of a qubit, with a parameter λ\lambda between 00 and 11, and the phase flip channel, which leaves the qubit alone with some probability and otherwise applies the Pauli operator ZZ, are each given by a pair of operation elements, and the two pairs look quite different. The exercise asks for a 2×22 \times 2 unitary matrix uu that turns one pair into the other, the phase flip elements being linear combinations of the phase damping elements with coefficients from the rows of uu, once the probability of no flip is chosen as a suitable function of λ\lambda.

A Qubit

The state space HH of the qubit is a two-dimensional complex Hilbert space with a fixed orthonormal basis |0⟩,|1⟩|0\rangle, |1\rangle. For i=0,1i = 0, 1 the projector PiP_i on HH is the rank-one operator Pi=|i⟩⟨i|P_i = |i\rangle\langle i| onto the basis vector |i⟩|i\rangle.✦

Two Pairs of Operation Elements

The phase damping elements with a real parameter λ\lambda are the ordered pair of operators on HH (Ek)k=0,1=(E0,E1),(E_k)_{k=0,1} = (E_0, E_1),(1) where, with matrices in the basis |0⟩,|1⟩|0\rangle, |1\rangle, E0=P0+1−λP1=(1001−λ),E_0 = P_0 + \sqrt{1-\lambda}\,P_1 = \begin{pmatrix} 1 & 0 \\ 0 & \sqrt{1-\lambda} \end{pmatrix},(2) E1=λP1=(000λ).E_1 = \sqrt{\lambda}\,P_1 = \begin{pmatrix} 0 & 0 \\ 0 & \sqrt{\lambda} \end{pmatrix}.(3)

The no-flip probability α\alpha belonging to λ\lambda is the real number α=1+1−λ2.\alpha = \frac{1 + \sqrt{1-\lambda}}{2}.(4)✦

The phase flip elements belonging to λ\lambda are the ordered pair of operators on HH (Ẽk)k=0,1=(Ẽ0,Ẽ1),(\tilde E_k)_{k=0,1} = (\tilde E_0, \tilde E_1),(5) where, with α\alpha the no-flip probability (4), Ẽ0=αI,\tilde E_0 = \sqrt{\alpha}\, I,(6) Ẽ1=1−α(P0−P1)=1−αZ,\tilde E_1 = \sqrt{1-\alpha}\,(P_0 - P_1) = \sqrt{1-\alpha}\, Z,(7) with II the identity on HH and Z=P0−P1Z = P_0 - P_1 the Pauli operator in the basis |0⟩,|1⟩|0\rangle, |1\rangle.

The mixing matrix uu belonging to λ\lambda is the 2×22 \times 2 complex matrix u=(α1−α1−α−α),u = \begin{pmatrix} \sqrt{\alpha} & \sqrt{1-\alpha} \\ \sqrt{1-\alpha} & -\sqrt{\alpha} \end{pmatrix},(8)✦ with α\alpha again given by (4), and with rows and columns indexed by 00 and 11, like the operation elements.

The Claim

The matrix uu answers the exercise.

Theorem 1. Let λ≥0\lambda \ge 0. Then the mixing matrix uu is unitary, u†u=uu†=I.u^\dagger u = u\,u^\dagger = I.(9)✦ If moreover λ≤1\lambda \le 1, then for k=0,1k = 0, 1 Ẽk=∑j=01ukjEj.\tilde E_k = \sum_{j=0}^{1} u_{kj}\, E_j.(10)✦

What Is Checked

The theorem is stated over the phase damping elements (1) and the phase flip elements (5), two ordered pairs of operators on a qubit built from the projectors P0P_0 and P1P_1 of a fixed orthonormal basis, and over the mixing matrix uu of (8), whose entries depend on λ\lambda through the no-flip probability α\alpha. What the library proves about these constructions is the unitarity (9) of uu: the matrix that carries the one pair into the other is a unitary.