Presentation + Paper
2 March 2020 Neural network based photon counting of summed single photon receivers
Nicholas C. Lantz, Jennifer N. Downey, Brian E. Vyhnalek, Sarah A. Tedder
Author Affiliations +
Proceedings Volume 11272, Free-Space Laser Communications XXXII; 112720P (2020) https://doi.org/10.1117/12.2545892
Event: SPIE LASE, 2020, San Francisco, California, United States
Abstract
Optical communication links, operating in low photon flux conditions, rely on an array of single photon counting detectors to receive the signal. Due to the reset time of these detectors, many separate detectors must be used to receive a continuous signal. Photon counting, channel combining, channel alignment, and digitization of the detected signal can be complex and expensive due to the parallel hardware required for each channel. This issue is compounded as the system scales to greater numbers of detectors due to the amount of hardware required and alignment requirements between each channel. The purpose of this research is to examine a photon counting and channel combining method which allows for photon detection channels to be summed into a single signal before digitization, eliminating the need for parallelized hardware. This reduction in parallel hardware has the potential to reduce the cost and complexity of the system. In this paper, a single layer fully connected neural network architecture is explored as a possible solution for the photon counting of summed photon detection channels. The signal to noise ratio of the combined signal was lower than that of the individual channels and was inversely proportional to the root of the number of channels being summed. The neural network signal processing implementation produced signal gain when the symbol phase remains constant. This is most likely due to the network exploiting the modulation structure of the signal and possibly offsets the losses incurred during analog summation.
Conference Presentation
© (2020) COPYRIGHT Society of Photo-Optical Instrumentation Engineers (SPIE). Downloading of the abstract is permitted for personal use only.
Nicholas C. Lantz, Jennifer N. Downey, Brian E. Vyhnalek, and Sarah A. Tedder "Neural network based photon counting of summed single photon receivers", Proc. SPIE 11272, Free-Space Laser Communications XXXII, 112720P (2 March 2020); https://doi.org/10.1117/12.2545892
Advertisement
Advertisement
RIGHTS & PERMISSIONS
Get copyright permission  Get copyright permission on Copyright Marketplace
KEYWORDS
Signal detection

Signal to noise ratio

Signal processing

Sensors

Neural networks

Photodetectors

Photon counting

Back to Top