Boson Sampling with Gaussian input states: efficient scaling and certification

Raphael A. Abrahao,Arman Mansouri,Austin P. Lund

2023-06-30

Abstract:A universal quantum computer of moderate scale is not available yet, however intermediate models of quantum computation would still permit demonstrations of a quantum computational advantage over classical computing and could challenge the Extended Church-Turing Thesis. One of these models based on single photons interacting via linear optics is called Boson Sampling. Although Boson Sampling was demonstrated and the threshold to claim quantum computational advantage was achieved, the question on how to scale up Boson Sampling experiments remains. To make progress with this problem, here we present a practically achievable pathway to scale Boson Sampling experiments with current technologies by combining continuous-variables quantum information and temporal encoding. We propose the combination of switchable dual-homodyne and single-photon detections, the temporal loop technique and scattershot-based Boson Sampling. We detail the required assumptions for concluding computational hardness for this configuration. Furthermore, this particular combination of techniques permits an efficient scaling and certification of Boson Sampling, all in a single experimental setup.

Quantum Physics

What problem does this paper attempt to address?

The paper attempts to address the problem of **how to effectively scale Boson Sampling experiments**. Specifically, Boson Sampling is a quantum computing model based on the interaction of single photons through a linear optical network, which can demonstrate the advantage of quantum computers over classical computers in certain tasks. However, current Boson Sampling experiments face several challenges, including:

1. **Advances in classical algorithms**: Better classical algorithms have pushed the threshold for quantum computing advantage to a higher number of input single photons.
2. **Difficulty in preparing multiple single photons**: As the number of input photons increases, it becomes increasingly difficult to prepare multiple single photons.
3. **Photon loss in linear optical networks**: As the network scale expands, the problem of photon loss becomes more severe.

To address these challenges, the authors propose a method that combines continuous variable quantum information and time encoding to achieve effective scaling and verification of Boson Sampling experiments. This method not only simplifies the experimental resource requirements but also allows for efficient verification of Boson Sampling within the same experimental setup. Specifically, the authors propose the following points:

- **Using continuous variable quantum states and time encoding**: By encoding information in the quantum modes of light and utilizing time delay techniques, spatial modes can be converted into time modes, significantly reducing the number of required spatial modes.
- **Combining dual homodyne detection and single-photon detection**: By switching between different detection methods in the experiment, efficient verification of Boson Sampling can be performed without changing the experimental setup.
- **Assuming the input state is a Gaussian state**: By assuming the input state is a Gaussian state, the covariance matrix can be used to efficiently estimate and verify the output state, thus avoiding the need for extensive repeated measurements of each output state.

In summary, this paper aims to provide a practical solution to overcome the scaling challenges faced by current Boson Sampling experiments and ensure their effectiveness and reliability in large-scale experiments.