Boson sampling with self-generation of squeezing via interaction of photons and atoms

Sergey V. Tarasov,Vladimir V. Kocharovsky

2024-09-14

Abstract:We suggest a novel scheme for generating multimode squeezed states for the boson sampling implementation. The idea is to replace a commonly used linear interferometer by a multimode resonator containing a passive optical element consisting of two-level atoms dispersively interacting with photons and self-generating a squeezed compound state of both bosons -- photons and atoms. The suggested scheme does not need (a) on-demand external sources of photons in squeezed or Fock quantum states and (b) numerous interchannel couplers which introduce phase noise and losses that prevent scaling up the system and achieving quantum advantage. The idea is illustrated by a setup based on a Bose-Einstein-condensed gas confined in a multimode resonator, one of whose optical modes is in the classical coherent regime. The joint probability distribution of photon and/or noncondensed atom numbers is calculated via a matrix hafnian that, for certain parameters of the system, is hardly to be effectively calculated by classical computers. Such experiments are at reach via existing cavity-QED and cold-gas technology.

Quantum Physics,Quantum Gases

What problem does this paper attempt to address?

The problem that this paper attempts to solve is to achieve quantum advantage in Boson Sampling experiments, that is, to complete tasks that are difficult for classical computers to calculate effectively through quantum systems. Specifically, the paper proposes a new scheme to generate multimode squeezed states for Boson Sampling experiments to overcome several major limitations in existing schemes:

1. **No need for external photon sources**: Traditional schemes usually require a large number of external light sources to generate photons in squeezed or Fock states. These light sources are not only difficult to synchronize, but may also introduce phase noise and losses, affecting the scalability of the system. The new scheme automatically generates squeezed states by using the non - resonant interaction between two - level atoms and photons in a multimode cavity, avoiding the dependence on external light sources.

2. **Reduce the use of channel couplers**: Traditional Boson Sampling experiments require multiple channel couplers to mix input modes. These couplers will introduce phase noise and losses, hindering the expansion of the system. The new scheme reduces the need for these couplers by directly generating squeezed states in a multimode cavity.

3. **Improve the scalability of the system**: The new scheme generates squeezed states in a multimode cavity, reducing photon losses and improving the scalability of the system, thus making it more likely to achieve quantum advantage.

### Core innovation points of the paper

- **Generation of squeezed states in multimode cavities**: The paper proposes to fill a medium in a multimode cavity. This medium can introduce counter - rotating terms, thereby directly generating squeezed states in the multimode cavity. This medium can be a partially condensed cold two - level Bose - atom gas. The non - resonant interaction between these atoms and photons brings the required squeezing effect.

- **The role of counter - rotating terms**: Counter - rotating terms play a key role in the Hamiltonian. They are responsible for generating non - zero anomalous correlations. These anomalous correlations make the system have statistical characteristics that are difficult for classical computers to simulate under certain parameters. For example, when the system is in thermal equilibrium, even in the mean - field description, counter - rotating terms can lead to non - trivial matrix hafnians.

### Mathematical model

The paper describes in detail a multimode cavity - QED (cavity - quantum electrodynamics) system, which contains multiple high - Q photon modes and a partially condensed cold two - level Bose - atom gas. The Hamiltonian \(\hat{H}\) of the system can be expressed as:

\[
\hat{H}=\int \hat{\psi}^\dagger_a\left(-\frac{\hbar^2}{2m}\nabla^2 + V_{\text{tr}}(r)+\frac{\hbar|\Omega(r)|^2}{\Delta_a}+\frac{g_a}{2}\hat{\psi}^\dagger_a\hat{\psi}_a\right)\hat{\psi}_a d^3r
\]
\[
+\int \hat{\psi}^\dagger_a\frac{\hbar}{\Delta_a}\sum_\nu\left(G^*_\nu\Omega\hat{a}^\dagger_\nu + G_\nu\Omega^*\hat{a}_\nu\right)\hat{\psi}_a d^3r
\]
\[
+\int \hat{\psi}^\dagger_a\frac{\hbar}{\Delta_a}\sum_{\nu,\nu'}\left(G^*_\nu G_{\nu'}\hat{a}^\dagger_\nu\hat{a}_{\nu'}\right)\hat{\psi}_a d^3r-\sum_\nu\hbar\Delta_\nu\hat{a}^\dagger_\nu\hat{a}_\nu
\]

where:
- \(\hat{\psi}_a\) is the atomic - field operator.
- \(\hat{a}_\nu\) and \(\hat{a}_{\nu'}\) are photon - mode operators.