Neuromorphic event sensors cut FSOC beacon acquisition latency to microseconds in UAV and HAP tests

The arXiv preprint by Francesco Nardo reports direct processing of asynchronous brightness-change events from a neuromorphic sensor for pointing, acquisition and tracking of multiple optical beacons. Experiments used an experimental dataset replicating realistic signal-to-background ratios and platform motion, comparing event-driven algorithms on accuracy and processing latency against conventional cameras limited by frame rate and blur.

Event cameras avoid the latency-frame-rate trade-off of frame sensors, enabling continuous tracking during rapid angular changes typical in free-space links between high-altitude platforms and UAVs. This aligns with broader industry moves toward low-SWaP terminals for non-terrestrial networks, where conventional PAT systems struggle with jitter and field-of-view constraints.

The work remains a controlled validation on recorded data rather than on-sky closed-loop tests; scaling will require radiation-tolerant hardware and integration benchmarks against existing quad-cell or focal-plane trackers. Follow-on flight campaigns on stratospheric platforms would provide the missing end-to-end jitter and link-budget data.

Future deployments could shorten initial acquisition windows from tens of milliseconds to sub-millisecond, directly increasing network availability for mobile optical backhaul.