Prof. Shlomi Arnon

Optical wireless communication (OWC) using orbital angular momentum (OAM)-encoded structured light enables high-dimensional multiplexing for next generation free-space links. However, atmospheric turbulence introduces phase distortions, modal coupling, and intensity fluctuations that significantly degrade detection performance and increase bit-error rate (BER). In this work, we propose a physics-informed hybrid quantum–classical framework based on a weighted conjugate light field (W-CLF) and a quantum neural network (QNN) for robust decoding of turbulence-distorted OAM signals. The proposed W-CLF extends conventional conjugate light field processing by incorporating multi domain adaptive weighting, fractional-order enhancement, and multiscale feature extraction to emphasize turbulence resilient modal structures while suppressing stochastic distortions. The resulting representation is processed by a variational quantum circuit that performs compact feature embedding and measurement-based classification under limited training data. Numerical simulations and tabletop optical experiments demonstrate that the proposed framework achieves a minimum BER of 6.10 × 10⁻⁵, outperforming classical CLF-based and deep-learning receivers while requiring up to 50% fewer training samples. To provide physical insight into the role of quantum processing, we further perform latent-space, qubit-scaling, and parameter-efficiency analyses, showing that variational quantum evolution improves feature separability and enables competitive performance with substantially fewer trainable parameters than comparable classical models. These results highlight the potential of physics-guided quantum-enhanced representation learning for turbulence resilient optical communication systems.