AI-Assisted Next-Gen Outdoor Optical Networks: Camera Sensing for Monitoring and User Localization

We consider outdoor optical access points (OAPs), which driven by recent advances in metasurface technology have gained increasing attention. Despite offering high data rates and strong physical-layer security, real-world deployments remain vulnerable to misuse, motivating a dedicated monitoring layer. We propose a user positioning and monitoring system that infers locations from spatial intensity measurements on a photodetector (PD) array. Our hybrid framework combines an optics-informed forward model and sparse, model-based inversion with a lightweight data-driven calibration stage, achieving high accuracy at low computational cost while retaining interpretability, stability, and robustness to nonidealities and device-specific distortions. Under identical hardware and training conditions (both with 5 × 105 samples), the hybrid method attains consistently lower mean-squared error than a generic deep-learning baseline while using substantially less training time and compute. Accuracy improves with array resolution and saturates around 60×60-80×80, indicating a favorable accuracy-complexity trade-off for real-time deployment. The resulting position estimates can be cross-checked with real-time network logs to enable continuous monitoring, anomaly detection (e.g., potential eavesdropping), and access control in outdoor optical access networks.