Ability of entanglement and purity to help to detect systematic experimental errors

1. Problem Statement

In quantum information processing, accurate readout of quantum states via measurement is fundamental. However, experimental data is inevitably affected by two types of errors:

• Statistical Errors: Arise from finite sampling (finite number of repetitions). These can be suppressed by increasing the number of measurements.
• Systematic Errors: Arise from environmental influences, hardware imperfections, or misalignments (e.g., incorrect wave plate angles, detector bias). These errors persist regardless of the number of repetitions and can lead to fundamentally flawed conclusions, such as non-physical state estimates or incorrect fidelity assessments.

While methods exist to handle statistical errors (e.g., confidence regions), distinguishing them from systematic errors remains a challenge. Existing tools often require prior knowledge of the error model. The authors address the need for a direct, efficient, and model-independent method to detect systematic errors in quantum experiments, specifically investigating how quantum state properties like purity and entanglement influence this detection capability.

2. Methodology

The authors propose a method based on comparing two different quantum state estimators derived from the same experimental data:

1. Unbiased Estimator ($\rho_{LS}$): The Least-Squares estimator obtained via linear inversion of the measurement frequencies. This estimator is unbiased but can yield non-physical states (matrices with negative eigenvalues) if systematic errors are present or if statistical fluctuations are large.
2. Biased Estimator ($\rho_{B}$): A physical state estimator obtained by projecting $\rho_{LS}$ onto the set of valid density matrices (positive semidefinite, unit trace) using convex optimization (minimizing the Hilbert-Schmidt distance).

The Detection Metric:
The core of the method is the distance $D$ between these two estimators:
$$D = \|\rho_{LS} - \rho_{B}\|_2$$

• Logic: If only statistical errors exist, $\rho_{LS}$ will converge to a physical state as the number of measurements $N \to \infty$, making $D \to 0$. If a systematic error exists, $\rho_{LS}$ may remain far outside the physical state space even for large $N$, resulting in a non-zero $D$.
• Statistical Bound: The authors derive a rigorous bound using the Vector Bernstein Inequality to calculate the probability $\delta_{sta}$ that a distance $D > \tau$ is caused solely by statistical fluctuations. If the observed $D$ exceeds a threshold $\tau$ with high confidence ($1 - \delta_{sta}$), a systematic error is certified.

Theoretical Investigation:
The paper analytically and numerically investigates the minimal state purity ($p_{min}$) required to detect specific systematic errors (e.g., misalignment of measurement bases). They compare:

• Single-qubit states.
• Two-qubit separable (product) states.
• Two-qubit entangled states.

Experimental Implementation:

• Source: A semiconductor GaAs quantum dot embedded in a micromachined piezoelectric substrate (bullseye cavity).
• State Generation: Biexciton-exciton radiative cascade produces polarization-entangled photon pairs.
• Control: Strain tuning is used to vary the fine-structure splitting, thereby controlling the purity of the emitted two-photon state (ranging from 0.56 to 0.92).
• Error Injection: A systematic error is introduced by applying a constant angular offset ($\Delta$) to the quarter-wave plate in the measurement setup of one qubit.
• Measurement: Quantum State Tomography (QST) is performed using local Pauli measurements.

3. Key Contributions

• Direct Detection Protocol: A user-friendly protocol that requires no prior knowledge of the error source, relying only on the comparison between unbiased and biased estimators.
• Role of Entanglement: Theoretical proof that entanglement significantly enhances error detection. Specifically, certain systematic errors (like the exchange of $\sigma_Y$ and $\sigma_Z$ bases, equivalent to a partial transposition) are undetectable using separable (product) states but are clearly detectable using entangled states.
• Purity Thresholds:
  • For single-qubit systems, a minimum purity of 0.75 is required to detect a specific basis misalignment error.
  • For two-qubit systems, the required purity drops significantly (to ~0.33 for entangled states), meaning even "noisy" entangled states are highly sensitive to systematic errors.
• Experimental Validation: Successful demonstration of the method using a solid-state quantum dot source, showing that the distance $D$ scales with state purity and effectively identifies wave plate misalignments.

4. Results

• Theoretical Bounds:
  • The minimal necessary purity ($p_{min}$) for a single qubit to detect a $\sigma_Y \leftrightarrow \sigma_Z$ swap error is 0.75.
  • For two-qubit systems, $p_{min}$ is significantly lower (0.29–0.34).
  • Crucial Finding: For the specific error of swapping measurement bases (partial transposition), only entangled states can detect the error. Separable states fail to violate positivity constraints under this specific error, regardless of purity.
• Experimental Data:
  • The experiment varied the state purity from 0.56 to 0.92.
  • The distance $D$ increased with the magnitude of the angular offset $\Delta$ and was consistently higher for higher purity states.
  • For highly entangled states (high purity), the method detected systematic errors with a confidence level of >90% for distances $D \geq 0.25$.
  • The experimental results matched the theoretical simulations perfectly.

5. Significance

• Robust Quantum Characterization: This work provides a critical tool for the "quantum hardware" community. As quantum processors scale, distinguishing between random noise and systematic calibration errors is vital. This method offers a way to certify the integrity of experimental setups without needing to know the specific nature of the fault.
• Resource Efficiency: It demonstrates that entanglement is not just a resource for computation, but a resource for metrology and error detection. Using entangled probes allows for error detection in states with lower purity, which are often easier to generate in noisy intermediate-scale quantum (NISQ) devices.
• Scalability: The method is theoretically scalable to multi-qubit systems. While the optimization complexity grows, the fundamental principle of comparing biased and unbiased estimators remains applicable.
• Practical Impact: The ability to detect subtle alignment errors (like wave plate misalignments) in real-time helps improve the fidelity of quantum communication protocols and quantum state tomography, directly addressing a major bottleneck in the realization of reliable quantum networks.

In conclusion, the paper establishes that entanglement and high purity act as amplifiers for systematic errors, making them easier to detect through the divergence between physical and non-physical state estimators. This insight guides the design of more robust quantum experiments and error mitigation strategies.