Enhancing Detection of Satellite Laser Ranging Signal in Low Signal-to-Noise Ratio Conditions

Laser ranging stations encounter low signal-to-noise returns when acquiring more distant satellite targets at lower elevations and in poorer atmospheric conditions. Signal tracks from successful detections of the retro-reflected light returning from a satellite can be identified in repeated range residuals that will also contain background noise. A close look at different observing scenarios and methods used to extract the satellite returns from background noise, developed over time at lunar and satellite laser ranging stations, reveals similar binning, trend removal and threshold approaches tuned to the individual stations, considering so far only parallelogram shapes. To generalise this common signal identification problem at laser ranging stations, the distribution of data recorded during a satellite pass is described using Poisson statistics, accounting for both the signal and noise elements. A model is then outlined and tested against real observations. We propose a method for the reduction of noise for the first-photon arrival effect typical of single-photon avalanche diodes (SPAD).<br>The number of entries that fall inside a 2D bin is compared against a threshold, defined by the likelihood ratio, to decide if the observed mean rate is closer to the one of the noise or of the signal. For the optimization of the bin size and threshold, we use Receiver Operating Characteristic (ROC) curves to choose the values that maximize true signal detection with minimum false positives.<br>We consider a new geometrical pattern for the 2D binning of the observation domain. The new pattern consists of hexagons and an improvement of 20% in signal detection was achieved compared to using equivalent parallelogram bins.