dqsweep

This repository is designed to analyze the performance (fidelity and latency) of multiple distributed quantum experiments over a configurable quantum network. It uses quantum network simulators such as NetSquid and SquidASM.

The research poster below introduces some distributed quantum computing concepts used in this repository:

Overview

The repository explores the performance of several quantum distributed experiments by sweeping given parameters (such as depolarization probabilities, gate times, and qubit coherence times) from a quantum network and then assesses two metrics:

• Average Fidelity (%): Density matrix output compared to the density matrix expected.
• Average Simulation Time (ms): Simulation time to execute the experiment.

The experiments implemented in this repository include:

• Nonlocal CNOT Gate with Two Teleportations (nonlocal_cnot_teledata.py): Implementation of the distributed CNOT gate using two quantum teleportations.
• Nonlocal CNOT Gate with Telegate (nonlocal_cnot_telegate.py): Implementation of a distributed CNOT gate between Alice and Bob presented in the paper [1].
• Distributed Grover on Two Qubits (dgrover_2.py): Implementation of a distributed Grover on two qubits with an initial ping-pong quantum teleportation exchange where Alice initiates the state of Bob's qubit.
• Distributed Grover on N Qubits (dgrover.py): General implementation of a distributed Grover on n qubits with an oracle that marks the state |1...1>.
• Distributed Quantum Fourier Transform (DQFT) on Two Qubits (dqft_2.py): Implementation of a distributed QFT on two qubits.
• Ping-Pong Teleportation (pingpong.py): A bidirectional quantum teleportation where a qubit is sent back and forth between Alice and Bob.
• Nonlocal Toffoli Gate (nonlocal_toffoli.py): Implementation of a distributed Toffoli gate between Alice, Bob, and Charlie.

Installation

1. Create a Python Virtual Environment:

python3 -m venv .venv
source .venv/bin/activate
python3 -m pip install -U pip

2. Install SquidASM (The credentials from Netsquid are required)

export NETSQUIDPYPI_USER=user1234
export NETSQUIDPYPI_PWD=password1234
git clone git@github.com:QuTech-Delft/squidasm.git
make install -C squidasm

3. Install the requirements

pip install -r requirements.txt

4. Verify the installation

pytest

Usage

The experiments are executed through the run_simulation.py script, which performs parameter sweeps in parallel, runs the specified distributed quantum experiment, and generates the results in a folder containing various files. To run the simulation:

1. Configure A Quantum Network:

• (--config): Provide a valid quantum network configuration or use the ones already provided (e.g., perfect.yaml, depolarise_link.yaml) in the configurations/ folder. The default configuration is perfect.yaml.

2. Setup the Experiment and Multiple Parameters:

• (--experiment): Choose the experiment to run (options include cnot_teledata, cnot_telegate, dgrover, dgrover2, dqft2, pingpong, and toffoli ). The default experiment is cnot_telegate.
• (--epr_rounds): Specify the number of EPR rounds per simulation. The default number is 10.
• (--num_experiments): Specify the number of simulation runs per parameter combination. The default number is 10.
• (--sweep_params): Define the comma-separated list of parameters to sweep (e.g., single_qubit_gate_depolar_prob, two_qubit_gate_depolar_prob, T1, T2, etc.). The default parameters are single_qubit_gate_depolar_prob,two_qubit_gate_depolar_prob.
• (--ranges): Provide for each swept parameter a valid range in the format "start,end,points". The default ranges are "0.0,0.8,10" "0.0,0.8,10".
• (--output_dir): Define the path of the directory to save the results. The default folder is results.

3. Execute the Simulation:

• Run the run_simulation.py script with the setup. The basic command to launch the simulation with the default configuration, experiment, and parameters (see below for a complete command line):

  python -m experiments.run_simulation

• The script creates an output subdirectory (named based on the experiment and swept parameters) that contains:
  • A CSV file with raw simulation results.
  • A TXT file with parameter-to-metric correlation values.
  • Heat map plots with the performance results.

The Distributed Experiments

The provided experiments:

• cnot_teledata: Nonlocal CNOT gate using two quantum teleportations.
• cnot_telegate: Nonlocal CNOT gate using telegate.
• dgrover: General Distributed Grover on n qubits that searches |1...1>.
• dgrover2: Distributed Grover on two qubits with an initial ping-pong exchange between Alice and Bob.
• dqft2: Distributed Quantum Fourier Transform on two qubits.
• pingpong: Ping-pong quantum teleportation between Alice and Bob.
• toffoli: Nonlocal Toffoli gate using two ebits and four bits in each direction.

Example Command 1: Nonlocal CNOT (Depolarization Sweep)

python -m experiments.run_simulation \
  --config configurations/perfect.yaml \
  --experiment cnot \
  --epr_rounds 10 \
  --num_experiments 100 \
  --sweep_params single_qubit_gate_depolar_prob,two_qubit_gate_depolar_prob \
  --ranges "0.0,0.8,10" "0.0,0.8,10" \
  --output_dir results

Example Result (Heat map) from command 1: Nonlocal CNOT (Depolarization Sweep)

Example Command 2: Distributed Grover on two qubits (Depolarization Sweep)

python -m experiments.run_simulation \
  --config configurations/perfect.yaml \
  --experiment dgrover2 \
  --epr_rounds 10 \
  --num_experiments 100 \
  --sweep_params single_qubit_gate_depolar_prob,two_qubit_gate_depolar_prob \
  --ranges "0.0,0.8,10" "0.0,0.8,10" \
  --output_dir results

Example Result (Heat map) from command 2: Distributed Grover on two qubits (Depolarization Sweep)

References

[1] Eisert, Jens & Jacobs, Karel & Papadopoulos, Periklis & Plenio, M. (2000). Optimal local implementation of nonlocal quantum gates. Phys. Rev. A. 62. 10.1103/PhysRevA.62.052317.