Lean, informalized

Nielsen & Chuang · Chapter 8, Quantum noise and quantum operations · p. 386

Exercise 8.31 - Averaging a phase kick over a normally distributed angle, one matrix element at a time

Random phase kicks damp each coherence by a Gaussian factor.

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

A system has the orthonormal basis |0⟩,|1⟩,|2⟩,…|0\rangle, |1\rangle, |2\rangle, \dots, and its state ρ\rho has the matrix elements ρnm=⟨n|ρ|m⟩\rho_{nm} = \langle n|\rho|m\rangle. A phase kick through the angle θ\theta is the unitary D(θ)D(\theta) with D(θ)|n⟩=e−inθ|n⟩D(\theta)|n\rangle = e^{-in\theta}|n\rangle. The system receives one such kick, with θ\theta drawn at random from a normal distribution of mean zero and variance 2λ2\lambda. The exercise is to show that the average over the kick multiplies each matrix element ρnm\rho_{nm} by e−λ(n−m)2e^{-\lambda(n-m)^2}.✦

The Kick and Its Average

The phase kicked element of a complex matrix ρ=(ρnm)n,m≥0\rho = (\rho_{nm})_{n,m \ge 0} at the angle θ\theta and the position (n,m)(n, m) is the complex number ⟨n|D(θ)ρD(θ)†|m⟩=e−inθρnmeimθ.\langle n|D(\theta)\,\rho\,D(\theta)^\dagger|m\rangle = e^{-in\theta}\,\rho_{nm}\,e^{im\theta}.(1)✦

The phase damped element ⟨n|ℰ(ρ)|m⟩\langle n|\mathcal{E}(\rho)|m\rangle of ρ\rho is the average of the kicked element over an angle with the normal distribution of mean zero and variance 2λ2\lambda: ⟨n|ℰ(ρ)|m⟩=14πλ∫−∞∞⟨n|D(θ)ρD(θ)†|m⟩e−θ2/4λdθ.\langle n|\mathcal{E}(\rho)|m\rangle = \frac{1}{\sqrt{4\pi\lambda}} \int_{-\infty}^{\infty} \langle n|D(\theta)\,\rho\,D(\theta)^\dagger|m\rangle\, e^{-\theta^2/4\lambda}\,d\theta .(2)✦

The decay factor γλ(k)\gamma_\lambda(k) at the separation kk, for real numbers λ\lambda and kk, is the real number γλ(k)=e−λk2.\gamma_\lambda(k) = e^{-\lambda k^2}.(3)✦

The Claim

The average (2) is ρnm\rho_{nm} times the Fourier transform of the normal density at n−mn - m, and the Fourier transform of a Gaussian is again a Gaussian.

Theorem 1. Let λ>0\lambda > 0, let ρ\rho be a complex matrix with entries ρnm\rho_{nm} for n,m=0,1,2,…n, m = 0, 1, 2, \dots, and let nn and mm be nonnegative integers. Then ⟨n|ℰ(ρ)|m⟩=γλ(n−m)ρnm.\langle n|\mathcal{E}(\rho)|m\rangle = \gamma_\lambda(n - m)\,\rho_{nm}.(4)✦

What Is Checked

The claim is stated over the phase damped element (2), the Gaussian average of the phase kicked element (1), and over the decay factor (3).

About these constructions the library proves (4), which evaluates the Gaussian integral (2) in closed form for every λ>0\lambda > 0.