3

I read on Wikipedia two different descriptions of the "Husimi-Q representation." One is that it is the Wigner function convolved with a Gaussian, which in particular results in a positive definite function. The other is that it is "essentially" (their words) the density matrix put into normal order. I had some trouble understanding why these are the same.

For instance, if we let $H=\omega a^\dagger a$, then the thermal state at inverse temperature $\beta$ is $$\rho=N\exp(-\beta H),$$ which (if I did everything right) normal-orders to $$:\rho:=N\exp(-\beta'H),$$ where $$\beta'=-\log(1-\beta\omega)/\omega.$$ This seems fairly pathological to me (ignoring the issue that this new density matrix doesn't seem to be normalized): we have $\beta'>\beta$, so the system is at a colder temperature, so I would expect the distribution in phase space to be "less blurry" rather than "more blurry," (certainly classically this is true), and at $\beta\omega=1$ we have singular behavior: for any temperature colder than $\omega$, it seems like the object we'll get out will assign negative probabilities to certain states.

Did I do the computation wrong? Does normal ordering here mean something different than pushing $a$'s to the right of $a^\dagger$'s? Are there other contexts in which we can think of normal ordering as smearing out distribution functions?

glS
  • 15,488
  • 5
  • 42
  • 114

2 Answers2

4

Q-function and P-function

The Husimi Q-function of a density matrix $\rho$ is defined by $$ Q_\rho(\alpha) = \frac{1}{\pi}\langle \alpha \vert \rho \vert \alpha \rangle$$ where $$\rho = \frac{1}{\pi}\int\rho_\text{coh}(\alpha,\alpha^\ast)\lvert \alpha \rangle\langle\alpha\lvert\mathrm{d}\alpha\mathrm{d}\alpha^\ast$$ in the coherent states. $P_\rho(\alpha,\alpha^\ast) := \frac{1}{\pi}\rho_\text{coh}(\alpha,\alpha^\ast)$ is the Glauber–Sudarshan P-function. It follows with $\langle \beta\vert \alpha\rangle = \mathrm{e}^{-\beta^\ast\beta/2-\alpha^\ast\alpha/2 + \beta^\ast\alpha}$ that \begin{align} Q_\rho(\beta) & = \int P_\rho(\alpha,\alpha^\ast)\langle\beta\vert \lvert \alpha \rangle\langle\alpha\lvert\vert \beta\rangle\mathrm{d}\alpha\mathrm{d}\alpha^\ast\\ & = \int P_\rho(\alpha,\alpha^\ast)\mathrm{e}^{-\beta^\ast\beta-\alpha^\ast\alpha + \beta^\ast\alpha + \beta\alpha^\ast}\mathrm{d}\alpha\mathrm{d} \alpha^\ast \\ & = \int P_\rho(\alpha,\alpha^\ast)\mathrm{e}^{-\lvert \beta-\alpha\rvert^2}\mathrm{d}\alpha\mathrm{d} \alpha^\ast \end{align}

Normal ordering and anti-normal ordering

The Q-function naturally normal orders $\rho$. Since $a\lvert \alpha \rangle = \alpha \lvert\alpha\rangle$, we have that $f(a)\lvert \alpha\rangle = f(\alpha)\lvert \alpha\rangle$ and $\langle \alpha \rvert f(a^\dagger) = \langle\alpha\rvert \alpha^\ast$, so for a normal-ordered symbol $f_N(a,a^\dagger)$ with all annihilators to the right and all creators to the left, we have $\langle \alpha\rvert f_N(a,a^\dagger)\lvert \alpha\rangle = f_N(\alpha,\alpha^\dagger)$, and so $$ Q_\rho(\alpha) = \frac{1}{\pi}\rho_N(\alpha,\alpha^\ast)$$

The P-function naturally anti-normal orders $\rho$. Expand $$ \rho_A(a,a^\dagger) = \sum_{i,j}\rho_{i,j}a^i (a^\dagger)^j$$ and insert the completeness relation $\mathbf{1} = \frac{1}{\pi}\int\lvert\alpha\rangle\langle\alpha\rvert\mathrm{d}\alpha\mathrm{d}\alpha^\ast$ to get $$ P_\rho(\alpha,\alpha^\ast) = \frac{1}{\pi}\rho_A(\alpha,\alpha^\ast)$$

What is a bit confusing is that this ordering prescription is exactly opposite to what it does on observables. One finds that anti-normal ordered expectation values are computed with the Q-function and normal ordered expectation values are computed with the P-function, i.e. \begin{align} \langle \mathcal{O}_A(a,a^\dagger) \rangle & = \int Q(\alpha,\alpha^\ast) \mathcal{O}_A(\alpha,\alpha^\ast)\mathrm{d}\alpha\mathrm{d}\alpha^\ast \\ \langle \mathcal{O}_N(a,a^\dagger) \rangle & = \int P(\alpha,\alpha^\ast) \mathcal{O}_N(\alpha,\alpha^\ast)\mathrm{d}\alpha\mathrm{d}\alpha^\ast \end{align}

ACuriousMind
  • 132,081
0

Preliminaries

  • $Q$ function and normal ordering — If an operator $B$ (which may or may not be a state) can be put in normal order, write it as $B_N(a,a^\dagger)\equiv\sum_{ij} c_{ij} a^{\dagger i}a^j$, then its $Q$ function is simply $$Q_B(\alpha)=\frac1\pi B_N(\alpha,\bar\alpha)=\frac1\pi\sum_{ij} c_{ij} \bar\alpha^i \alpha^j.$$ In other words, you get the expression for $Q$ by simply replacing $a\to\alpha$ and $a^\dagger\to\bar\alpha$. This is easy to prove, as $\langle \alpha|B_N(a,a^\dagger)|\beta\rangle=B_N(\beta,\bar\alpha)$.

    Note that here I said "$B$ can be put in normal order" and not "$B$ is normally ordered". This might seem like a subtlety but I think it helps to distinguish these concepts: $B$, as an operator, is neither normally nor antinormally ordered. The concepts of "ordering" only make sense when you refer to specific ways to write the operators, before interpreting $a,a^\dagger$ as actual operators acting on some space. See What does "not applying the CCR" mean exactly? for more details on this point.

  • $P$ function and antinormal ordering — Similarly if an operator $B$ can be put in antinormal order, write it as $B_{A}(a,a^\dagger)$, then $P_B(\alpha)=\frac1\pi B_{A}(\alpha,\bar\alpha)$.

  • Convolution relation between $P$ and $W$ — The standard way to introduce $P$ and $Q$ functions is explicitly via the expression $Q_\rho(\alpha) = \frac1\pi \langle\alpha|\rho|\alpha\rangle$, and implicitly via $$\rho = \int \mathrm d^2\alpha P_\rho(\alpha) |\alpha\rangle\!\langle\alpha|.$$ These two are sufficient to get the expression with the convolution, as already shown in the other answer: $$Q_\rho(\alpha) = \frac1\pi \langle\alpha|\rho|\alpha\rangle = \frac1\pi\int \mathrm d^2\beta P_\rho(\beta) \lvert\langle\alpha|\beta\rangle\rvert^2 = \frac1\pi\int \mathrm d^2\beta P_\rho(\beta) e^{-|\alpha-\beta|^2}.$$ A similar way to relate $W$ and $Q$ via convolution with a Gaussian is also discussed in Show that the Husimi Q function equals $Q(\alpha)=\frac 1 \pi\langle \alpha |\hat \rho|\alpha\rangle$ from its relation with the Wigner function.

Some explicit examples might be useful to better understand this.

Example 1: $B=a^\dagger a$

Consider the operator $B\equiv a^\dagger a$. This is already written in normal order, $B_N(a,a^\dagger)=a^\dagger a$, and thus $$Q_B(\alpha) = \frac1\pi |\alpha|^2.$$ Writing it in antinormal order, we instead get $B_{A}(a,a^\dagger)= aa^\dagger-1$, and thus $$P_B(\alpha) = \frac1\pi (|\alpha|^2-1).$$ For consistency, we can check that we have the integral representation: $$a^\dagger a = \int \mathrm d^2\alpha \left(\frac{|\alpha|^2-1}{\pi}\right) |\alpha\rangle\!\langle\alpha|,$$ which amounts to the identity $$ n \delta_{n,m} = \frac1\pi\int \mathrm d^2\alpha (|\alpha|^2-1) e^{-|\alpha|^2} \frac{\alpha^n\bar\alpha^m}{\sqrt{n!m!}}. $$ Proving this identity is easy enough switching to polar coordinates, and is essentially equivalent to what's used in this answer. We can also check how $Q$ is correctly reproduced via convolution of $P$ with a Gaussian: $$\frac1\pi \int \mathrm d^2\beta P_B(\beta) e^{-|\alpha-\beta|^2} = \frac1{\pi^2}\int \mathrm d^2\beta (|\beta|^2-1)e^{-|\alpha-\beta|^2} = \frac{1}{\pi}|\alpha|^2.$$

Example 2: $B=a$

Perhaps an even simple example is $B=a$. In this case normal and antinormal orderings coincide: $B_A(a,a^\dagger)=B_N(a,a^\dagger)=a$, and thus so do $P$ and $Q$ representations: $$P_B(\alpha) = Q_B(\alpha) = \frac1\pi \alpha.$$ You can then also again check for consistency that the identities $$a = \int \mathrm d^2\alpha \left(\frac{1}{\pi}\alpha\right)|\alpha\rangle\!\langle\alpha|, \\ \frac1\pi\alpha = \frac1\pi\int \mathrm d^2\beta \left(\frac1\pi \beta\right)e^{-|\alpha-\beta|^2}$$ are satisfied (the first one is the representation of $a$ via the $P$ function, and the second one is the representation of $Q$ via convolution of a Gaussian with $P$).

Example 3: $\rho=|0\rangle\!\langle 0|$

Let's consider an example that's less trivial: the vacuum state, $\rho=|0\rangle\!\langle 0|$. In this case, it's less trivial to find normal/antinormal orderings for $\rho$, but they still exist. For example, a tidy expression for the normal ordering, discussed in Relationship between normal-ordered vacuum state and parity operator, is $|0\rangle\!\langle 0| = N(e^{-a^\dagger a})$, which in other words means $$\rho_N(a,a^\dagger) = \sum_{k=0}^\infty \frac{(-1)^k}{k!} a^{\dagger k}a^k.$$ In fact, as shown here, we have the integral representations $$ |0⟩\!⟨0| = \frac1\pi\int d^2\gamma e^{-|\gamma|^2/2}D(\gamma) = \frac1\pi\int d^2\gamma e^{-\bar\gamma a} e^{\gamma a^\dagger} = \frac1\pi\int d^2\gamma e^{-|\gamma|^2} e^{\gamma a^\dagger} e^{-\bar\gamma a}, $$ where $D(\gamma)\equiv e^{\gamma a^\dagger-\bar\gamma a}$ is the displacement operator, and I juggled with its standard identities to get the different expressions. Note these expressions are symmetrically, antinormally, and normally ordered, respectively, and thus immediately also provide us with the $P$ and $Q$ representations for the vacuum state: $$ P_\rho(\alpha) = \frac1{\pi^2} \int \mathrm d^2\gamma e^{-\bar\gamma \alpha}e^{\gamma\bar\alpha} = \delta^2(\alpha), $$ $$Q_\rho(\alpha) = \frac1{\pi^2} \int \mathrm d^2\gamma e^{-|\gamma|^2} e^{\gamma\bar\alpha}e^{-\bar\gamma \alpha} = \frac1\pi e^{|\alpha|^2}.$$ Of course, these expressions coincide with the well-known ones that can be derived pretty much directly from the definitions of $Q$ and $P$ given at the beginning of the post.

glS
  • 15,488
  • 5
  • 42
  • 114