Noise simulation: Pauli channels, amplitude damping and readout error

Ouverte
  • Quelque chose ne va pas ?
#8 0 commentaires 0 réactions 0 personnes assignées Voir sur GitHub

Personne n'a encore pris cette issue.

Évaluation

Difficulté
4/5
Temps estimé
3-5 jours
Accessibilité débutants
55/100
Type d'issue
Fonctionnalité
Clarté
Clairement spécifiée
Activité
Active
Stack technique
python, rust
Domaine
backend

Piste de recherche

Start with the per-shot dynamic branch in src/circuit.rs and the measurement and renormalisation pattern in src/backend.rs; trace how apply_gate_kernel and probability_of_one are used. Implementing this spans gate execution, noise sampling, and fast-path behavior. Done means run supports the listed noise channels and the state inspection methods reject gate noise as specified.

Rédigé par le modèle d'indexation à partir du texte de l'issue.

Description

enhancement
What you want to do
Summary

Every result today comes from an ideal, noiseless device. Studying error mitigation, error correction or how an algorithm holds up on real hardware needs gate noise and readout error. Right now users have to switch to another simulator for that.

Implementation notes
  • Use quantum trajectories (Monte Carlo wavefunction). The path that simulates each shot separately
    (QuantumCircuit::run, dynamic branch in src/circuit.rs) already does what this needs: one state
    per shot, a random generator, and measure_qubit / reset_qubit.
  • Pauli channels: after a gate, draw a random Pauli with the channel's probabilities and apply it with
    apply_gate_kernel. Single-qubit gate fusion has to stop at every noise point, or fuse only
    between noise points.
  • Amplitude damping is not a mixture of unitaries. Its Kraus branch probabilities depend on the state,
    p(jump) = γ · P(qubit = 1), and probability_of_one already computes that. Then apply the
    non-unitary 2×2 matrix and renormalise, the same pattern as measure_qubit in src/backend.rs.
  • Readout error only flips classical bits after sampling, so circuits that only measure at the end
    keep the fast path when readout is their only noise.
  • statevector(), probabilities() and expectation() should raise when given gate noise,
    since the state is a mixture. A density-matrix backend could be a later issue.
What it could look like
noise = qsengine.NoiseModel(
    depolarizing=1e-3,        # after every single-qubit gate
    depolarizing_2q=1e-2,     # after every controlled gate, on each qubit it touches
    amplitude_damping=5e-4,   # T1-style decay after every gate
    readout=0.02,             # chance each measured bit is flipped
)
counts = qc.run(shots=4000, seed=1, noise=noise)
Contributing
  • I would like to open a pull request for this (see CONTRIBUTING.md).
Langage dominant
Rust
Étoiles
1
Forks
0
Aucune PR mergée en 30 j

Préparer son environnement

Par où commencer

  1. Lisez l'issue en entier, puis le guide de contribution du projet.
  2. Signalez en commentaire que vous la prenez — cela évite que deux personnes fassent le même travail.
  3. Forkez le dépôt et travaillez sur une branche.
  4. Ouvrez une pull request qui référence le numéro de l'issue.

Autres issues de kaiwave/qsengine

Toutes les issues de kaiwave/qsengine

Issues similaires

Plus d'issues Rust

Recevez les nouvelles issues par e-mail

Un résumé court des issues GitHub adaptées aux débutants.