Modeling the Effects of Gaseous Absorption and Cloud Attenuation for V-band using Deep Learning

• Published in: Digest - IEEE Antennas and Propagation Society. International Symposium (1995) pp. 1589 - 1590
• Main Authors: Gesner, R. L., Gupta, A., Argyres, J., Christodoulou, C. G., Lane, S., Murrell, D., Hong, E., Tarasenko, N.
• Format: Conference Proceeding
• Language: English
• Published: IEEE 05.07.2020
• Subjects: Atmospheric losses; Atmospheric modeling; Attenuation; Cloud attenuation; Cloud computing; Deep learning; Machine Learning; Millimeter wave propagation; Neural networks; Predictive models; Terrestrial atmosphere
• ISSN: 1947-1491
• DOI: 10.1109/IEEECONF35879.2020.9329643

A deep neural network is proposed to study the attenuating effects of the Earth's atmosphere on the W/V-bands of the millimeter wavelength portion of the spectrum. Validation of atmospheric propagation models in the W/V-bands has become an increasingly important subject of communications research, but presents significant hurdles when testing these models due to the great variability in the atmospheric parameters that influence propagation attenuation. The effects of molecular gas resonances and hydrosols become very pronounced due to the short wavelength in these bands. This research employs a multilayered deep learning model to learn and predict the attenuation at V-band frequencies. The data is collected from the ongoing W/V-band Terrestrial Link Experiment (WTLE) in Albuquerque, NM. WTLE uses weather sensors and the received power data across the link to study the effects of atmosphere on V-band propagation.