bosonic-grid-states-qec

name

bosonic-grid-states-qec

description

Bosonic quantum error correction using Gottesman-Kitaev-Preskill (GKP) grid states and programmable nonlinear bosonic circuits. Covers grid state preparation, logical qubit encoding in continuous-variable oscillators, Wigner function characterization, and fault-tolerant bosonic QEC architectures. Use when working on: bosonic codes, GKP states, cat/qubit encodings, continuous-variable QEC, nonlinear bosonic gates, or oscillator-based quantum computing. Triggers: bosonic QEC, GKP grid states, cat codes, binomial codes, continuous-variable quantum computing, oscillator qubit, bosonic circuits.

Bosonic Grid States & Programmable Nonlinear Bosonic Circuits for QEC

Overview

Bosonic quantum error correction encodes logical qubits into infinite-dimensional harmonic oscillators using discrete grid states (GKP), cat states, or binomial states. The key insight: a single physical mode with structured encoding can detect and correct errors without the overhead of multi-qubit repetition codes.

Core Concepts

GKP Grid States

GKP states encode logical information in the periodic structure of position (q) and momentum (p) quadratures:

|μ,ν⟩_GKP ∝ Σ_n |q = √π(2n+μ)⟩ · e^{i√π ν p}

• Logical |0⟩ and |1⟩ are displaced combs in phase space
• Grid spacing √π provides correctability against small shifts
• Errors appear as displacements: shift < √π/2 → correctable

Error Model

Bosonic errors map to phase-space displacements:

• Shift errors: δq, δp displacements in quadrature space
• Photon loss: amplitude damping → biased shift errors
• Dephasing: random displacement in p-direction

Correction via Steane-type syndrome measurement: modular quadrature measurement reveals shift without collapsing logical information.

Programmable Nonlinear Bosonic Circuits

Recent work (arXiv:2604.21824) demonstrates:

• Engineered Kerr nonlinearities for universal bosonic gates
• Programmable beam-splitter operations via parametric coupling
• SNAP (Selective Number-dependent Arbitrary Phase) gates for Fock-state control
• Cross-Kerr interactions for entangling logical bosonic modes

Methodology

Step 1: Choose Bosonic Code

Code	Encoding	Best For
GKP	Periodic grid in phase space	General shift errors
Cat	Superposition of coherent states	Photon loss channels
Binomial	Number-state superpositions	Targeted loss+dephasing
Pair-cat	Two-mode entangled states	Biased noise

Step 2: State Preparation

GKP state preparation approaches:

1. Circuit-based: Oscillator-reset + sequence of conditional displacements
2. Measurement-based: Heralded preparation via ancilla coupling
3. Autonomous: Engineered dissipation stabilizing the grid manifold

Step 3: Error Syndrome Extraction

1. Apply controlled-displacement with ancilla qubit
2. Measure ancilla → reveals mod(q, √π) or mod(p, √π)
3. Conditional displacement correction based on syndrome
4. Repeat for both quadratures

Step 4: Logical Gate Implementation

Gate	Physical Implementation
X_L	Displacement by √π in q
Z_L	Displacement by √π in p
H_L	Phase-space rotation (π/4 shear)
CNOT	Entangling cross-Kerr + displacements
T_L	Magic state injection + distillation

Architecture Patterns

Single-Oscillator QEC Cycle

Oscillator → Syndrome Extraction → Classical Decoding → Conditional Correction
     ↑                                                        |
     └────────────────────────────────────────────────────────┘

Multi-Mode Architecture

[Mode 1] ──┬── [BS gate] ──┬── [Mode 2]
            │                │
        [SNAP]          [SNAP]
            │                │
        [Kerr]          [Kerr]
            └── [XK] ────────┘

Implementation Guidance

Wigner Function Analysis

# Wigner function for GKP state characterization
import numpy as np

def wigner_gkp(q_grid, p_grid, mu=0, nu=0, delta=0.3, n_peaks=5):
    """Approximate Wigner function of finite-energy GKP state.
    
    Args:
        q_grid, p_grid: Phase space grids
        mu, nu: Logical encoding parameters (0 or 1)
        delta: Gaussian envelope width (finite energy)
        n_peaks: Number of grid peaks to include
    """
    W = np.zeros_like(q_grid)
    for n in range(-n_peaks, n_peaks+1):
        q0 = np.sqrt(np.pi) * (2*n + mu)
        envelope = np.exp(-(q_grid - q0)**2 / (2*delta**2))
        W += envelope * np.cos(np.sqrt(np.pi) * nu * p_grid)
    return W

Syndrome Decoding

def gkp_syndrome_decode(syndrome_q, syndrome_p, sqrt_pi=np.sqrt(np.pi)):
    """Decode GKP shift syndrome to correction displacement.
    
    Returns correction (dq, dp) to apply.
    """
    dq = syndrome_q % sqrt_pi
    dp = syndrome_p % sqrt_pi
    
    # Map to [-sqrt_pi/2, sqrt_pi/2)
    if dq > sqrt_pi / 2:
        dq -= sqrt_pi
    if dp > sqrt_pi / 2:
        dp -= sqrt_pi
    
    return -dq, -dp

Key Metrics

Metric	GKP Target	Cat Target
Threshold displacement	√π/2 ≈ 0.886	Depends on α
Logical error rate	~exp(-πΔ²) per cycle	~exp(-2
Overhead vs rep code	~10x lower	~5x lower
Gate fidelity target	>99.9%	>99.5%

Error Budget

1. Finite-energy corrections: Δ² ∝ 1/⟨n⟩ — higher photon count = better approximation
2. Syndrome measurement errors: Ancilla infidelity propagates as logical error
3. Gate infidelity: Nonlinear gate errors accumulate per QEC cycle
4. Thermal noise: Excitation drift between correction cycles

Hardware Platforms

Platform	Oscillator Mode	Nonlinearity Source
Superconducting	3D/transmon cavity	JRM modulator, SNAIL
Trapped ions	Motional modes	Mølmer-Sørensen
Photonic	Temporal/freq modes	χ(2) crystals
Mechanical	MEMS resonators	Piezoelectric coupling

Pitfalls

• GKP states require high photon numbers; finite-energy envelope causes intrinsic error floor
• Cross-Kerr gates can be slow; gate times vs coherence time budget
• Syndrome measurement back-action on oscillator — use quantum non-demolition schemes
• Mode cross-talk in multi-oscillator architectures

References

• arXiv:2604.21824 — Programmable nonlinear bosonic circuits for grid states
• Gottesman, Kitaev, Preskill (2001) — Original GKP encoding
• Terhal et al. (2020) — Bosonic QEC survey
• Albert et al. (2018) — Performance bounds for bosonic codes