arXiv:2610.21606
Practical Fidelity Estimation for Noisy Quantum Devices Using Direct and Survival-Probability-Based Protocols
Ahmed Aman, Kaustav Chatterjee, Younes Javanmard, Tanmoy Pandit
Abstract
Reliable characterization of quantum states and quantum processes is central to the development of scalable quantum computing. Standard characterization techniques such as full quantum state tomography become rapidly impractical as system size increases. In this work we present two software-oriented fidelity estimation protocols that avoid full tomographic reconstruction while remaining experimentally relevant for noisy intermediate-scale quantum devices. The first protocol is direct fidelity estimation (DFE), where the fidelity between a target state and an experimentally prepared state is estimated from importance-sampled Pauli measurements. The second protocol is a survival-probability-based method that reconstructs fidelity from experimentally accessible moments of a noisy evolution and its approximate inverse, and is intrinsically robust against state-preparation-and-measurement (SPAM) errors. We discuss the theoretical basis of both approaches, analyze their noise sensitivity, and present numerical evidence on representative multiqubit states under depolarizing, dephasing, and amplitude-damping channels. For DFE we show that attenuation-aware correction improves robustness against measurement noise. For the survival-probability approach we show that an $L_2$-optimized expansion substantially outperforms naive Taylor truncation in high-noise regimes.
AI Summary
The paper examines two routes to estimating fidelity without full tomography: standard direct fidelity estimation (DFE) using importance-sampled Pauli measurements, and polynomial reconstruction of a quantity derived from forward–inverse evolution cycles. It provides implementation details, readout-noise corrections, and small-system numerical comparisons. For DFE, measured Pauli expectations are divided by calibrated attenuation factors. Independent per-qubit readout flips give attenuation $(1-2\alpha)^{w}$ for a weight-$w$ string, although the simulations instead use a single effective binary-outcome flip probability of $5\%$. Four-qubit simulations use $\epsilon=\delta=0.05$, corresponding to 8,000 sampled Pauli indices per trial, across several target states and noise channels. Corrected estimates are compared with directly calculated state overlaps. The second method approximates $F_{\mathcal Q}=\langle\langle\rho_0|\sqrt{\mathcal Q}|\rho_0\rangle\rangle$, where $\mathcal Q=\mathcal K^\dagger\mathcal K$, using moments of $\mathcal Q$. Polynomial coefficients minimize integrated squared error against $\sqrt{\lambda}$. Two-qubit simulations with degrees 2–4 and approximation interval $[0.05,1]$ show improved agreement over Taylor truncation at stronger noise, using exact moments and a matrix-square-root reference. The operational qualifications are substantial: measured return probabilities supply the required moments only when the implemented inverse equals $\mathcal K^\dagger$, which need not be trace-preserving. The proxy equals physical fidelity under the sufficient condition that the interaction-picture error map is positive semidefinite. Spectral containment must also be checked or enforced. The survival simulations do not test finite-shot noise, imperfect inverses, or SPAM robustness. DFE requires accessible target-state Pauli information, and attenuation correction can amplify sampling noise.
Central claim may not hold: The abstract claims intrinsic SPAM robustness although the survival estimator has no demonstrated general SPAM cancellation and its simulations include no SPAM fluctuations. The proposed experimental survival measurements yield the required adjoint-cycle moments only under a restrictive inverse-implementation condition that a generic approximate physical inverse does not satisfy. The survival reconstruction estimates an adjoint-cycle proxy rather than physical state fidelity in general, substantially narrowing the abstract's claim. The survival simulations use a fixed polynomial interval that excludes occupied eigenvalues of the tested channels at high noise, without reporting the rescaling or interval adjustment required by their stated assumptions.
Written by gpt-6.1-sol from the paper's full text. Automated: it can be wrong.
- Comments
- 9 pages, 3 figures