A library to help better understand Aaronson-Arkhipov Boson Sampling (AABS)

# Boson Sampling Library

The following library was developed to better understand Aaronson-Arkhipov Boson Sampling (AABS) (the acronym "AABS" for the particular Boson Sampling setup was encountered in Kruse et al.'s paper A detailed study of Gaussian Boson Sampling) by simulating the behavior of fock state inputs through a random unitary based linear interferometer and calculating the probabilities of certain fock state ouptuts.

## Installation

You can either

pip install bosonsampling


or clone the library and within the directory (as well as having your favorite python environment activate) run the following:

pip install .


## Usage

1. Import bosonsampling for core functionality and the random_interferometer() function from strawberryfields.utils to set up a random interferometer.

import bosonsampling as bs
from strawberryfields.utils import random_interferometer


# creates a 4x4 unitary matrix
random_interferometer = random_interferometer(4)

3. Create your input Fock states

# 0 photons in mode 1
# 1 photon in mode 2
# 3 photons in mode 3
# 4 photons in mode 4
photons_in = [0,1,3,4]

4. You can either:

a. Measure the probability of a certain output state

photons_out = [0,0,0,8]
output_probability = bs.output_probability(photons_in, photons_out, random_interferometer)


b. OR Measure the probability of ALL output states

for photon_out_configuration in bs.gen_output_configurations(sum(photons_in), len(photons_in)):

output_probability = bs.output_probability(photons_in, photon_out_configuration, random_interferometer)

print("Probability of photon output {0} : {1}".format(photon_out_configuration, output_probability))


## Example Files

Example files have been included to explain different possible usages with bosonsampling and can be found in the examples folder.

All files prefixed with bs_ are dedicated solely to bosonsampling.py.

However, there are files prefixed with sf_ that show how identical behavior can be obtained with Strawberry Fields (great for verifying results) and how bosonsampling can be used in conjunction with it as well.

• bs_single_photon_test.py - demonstrates a single photon entering the interferometer and calculating the probability of all possible outputs

• bs_multiple_photon_test.py - demonstrates multiple photons entering the interferometer and calculating the probability of all possible outputs, emphasizing how gen_output_configurations() can come in handy for such scenarios.

• bs_aabs_setup.py - demonstrates a setup similar to what would be called for in the Aaronson-Arkhipov paper with a multiple, single-photon fock state input with the modes right next to eachother and how the probability of detecting a "collision" (an output state that has multiple photons in a single mode) is incredibly low

• bs_aabs_scaling.py - Is essentially a duplicate of bs_aabs_setup.py but has an outer loop designed to make the interferometer larger with each iteration but keeps the number of inputted photons identical so one can observe the decreasing probability of collision.

• sf_single_photon_test.py - Is bs_single_photon_test.py but implemented with Xanadu's Strawberry Fields framework

• sf_multiple_photon_test.py - Is bs_multiple_photon_test.py but implemented with Xanadu's Strawberry Fields framework.

## Contributing

If you find a bug or have an idea to improve the library, please feel free to either make an Issue or a Pull Request with your suggested changes! If you are contributing code, please do note that this library attempts to follow the PEP-8 Style Guide for Python Code as well as using Google Style Python Docstrings

## Credits

The following resources were used to develop the main functions behind library:

### Videos

Quantum Supremacy via Boson Sampling: Theory and Practice | Quantum Colloquium by the Perimeter Institute with Scott Aaronson as the guest speaker.

Provides the formula in the first few minutes for calculating the probability of a certain photon output configuration (although the papers below also have equivalent formulations).

Quantum Complexity Theory: Lecture 9 - Boson Sampling by Sevag Gharibian

Provided an excellent explanation of the column and row indexing/copying required to generate a matrix, which is the basis for both the default and memory efficient implementations of the submatrix generator in the library.

Generating integer partitions by Jerome Kelleher

Provided the integer partitioning function which is used to help generate all possible output photon states in gen_output_configurations() function

Get all permutations of a numpy array by Bill Bell

Provided a method of using sympy to generate all unique permutations of a given numpy array (works with ordinary lists too) and is part of the gen_output_configurations() function

### Papers

An introduction to boson-sampling by Gard et al.

A detailed study of Gaussian Boson Sampling provided the origin of the AABS acronym used throughout the project and has a very nice illustration on page 2 that describes the column and row indexing necessary to generate the proper submatrix that Sevag Gharibian also goes over.

## Websites

Boson sampling and the permanent

provided the basis for code in sf_single_photon_test.py and sf_multiple_photon_test.py

These are resources that did not contribute directly to the code but were invaluable in learning more about boson sampling:

Experimental Boson Sampling by Tillman et al.

Introduction to boson-sampling by Peter Rohde

Gaussian Boson Sampling Complexity - Part 1 by Andrew Tanggara

Note: This video was barely consulted, but appred to be very well made and is given by Alex Arkhipov himself! Quantum computing with noninteracting particles - Alex Arkhipov

## Project details

Uploaded source
Uploaded py3