arXiv:2610.11001
Structured leakage in OAM-encoded qubits revealed by distributed quantum feature extraction
Yayu Mo, Nazifa Rumman, Sanjeev Shapkota, Thomas A. Searles, Sanjaya Lohani
Abstract
Orbital angular momentum (OAM) modes have emerged as an important platform for high-dimensional quantum communication but remain vulnerable to environmental noise and channel imperfections. Here we introduce a logical-qubit representation of OAM propagation in which modes evolve as qubits through noisy quantum channels. In this representation, turbulence manifests as leakage from the logical OAM subspace into turbulence-induced leakage components, providing a unified description of degradation in both single-photon and coherent-beam regimes. We show that turbulence-induced leakage exhibits systematic dependence on turbulence strength and OAM order, revealing structured signatures of information redistribution beyond the logical subspace. Leveraging this insight, we develop a distributed quantum feature-extraction scheme that encodes high-dimensional OAM states into quantum representations and employs quantum convolutional operations to extract turbulence-robust features. Simulations and experimental measurements demonstrate robust extraction of these features across a broad range of turbulence strengths, OAM orders, and propagation conditions. Our results establish a direct link between OAM-channel dynamics, structured leakage, and quantum learning, providing a physics-informed framework for characterizing turbulence-degraded high-dimensional optical states and extracting robust signatures of turbulence-induced information redistribution.
AI Summary
The paper asks how atmospheric turbulence redistributes information outside an OAM-encoded logical-qubit subspace, and whether signatures of that redistribution remain accessible in features extracted from distorted intensity images. Logical states are encoded in $|+\ell\rangle$ and $|-\ell\rangle$. Propagated fields are projected onto Laguerre–Gaussian modes to construct scattering matrices; their logical blocks define ensemble-averaged, completely positive, trace-decreasing maps, with trace loss interpreted as leakage. Split-step Kolmogorov-turbulence simulations use a $256\times256$ grid, five phase screens, 0.1 m propagation, and OAM orders $|\ell|=1$–6. Sampling up to 10,000 turbulence realizations yields leakage trends that increase with turbulence strength and OAM order; the paper reports an approximately linear dependence on turbulence strength. Simulated turbulence-derived channels are also compared with bit-flip, phase-flip, depolarizing, and amplitude-damping channels. For feature extraction, intensity images are divided into $3\times3$ patches, angle-encoded into nine-qubit circuits, and processed with fixed random quantum layers. Features are evaluated using SSIM and a geometric-mean fidelity of encoded subsystem states. Optical measurements use SLM-generated turbulence with both a 785 nm continuous-wave beam and an 810 nm SPDC photon-pair source. The photon-source measurements report SSIM above 0.6 and encoded-state fidelity above 0.8, with qualitative agreement between measured and simulated distortions. The stated objective is feature accessibility rather than computational advantage. Experimental limitations include low photon flux, background noise, and reduced signal-to-noise ratio, requiring adjusted encoding range. Extension to larger, higher-dimensional, and more realistic propagation scenarios is left open.
Central claim may not hold: The leakage formula squares coefficients already defined as populations and normalizes over a truncated basis, so it does not calculate the claimed unconditional probability of leaving the logical subspace. The reconstruction formula incorrectly equates a mode-diagonal mixture with a coherent pure-state projector, which generally has different coherence and purity. The proposed CPTP completion omits the ensemble normalization and can require the square root of a negative operator even for identity-channel realizations. The spatial-intensity formula uses global mode overlaps instead of position-dependent fields and therefore yields a spatial constant rather than an interference pattern. Image-feature similarities and idealized fidelities of encoded image patches do not establish quantitative recovery of physical leakage or fidelity of the propagated optical quantum state.
Written by gpt-6.1-sol from the paper's full text. Automated: it can be wrong.