rfroicnn_compressed.pdf (8.33 MB)

RFROI–CNN WSS – Radio Frequency Regions of Interests Convolutional Neural Network for Wideband Spectrum Sensing

Download (8.33 MB)
posted on 2022-06-23, 03:33 authored by Adam OlesińskiAdam Olesiński, Zbigniew Piotrowski

In this paper, we propose the application of deep convolutional neural networks for the detection of radio signals in the broadband frequency spectrum. The RFROI–CNN (Radio Frequency Regions of Interests) approach uses a different way of proposing regions than in typical convolutional object detectors in images – instead of proposing regions based on the distinguishing features of specific radio transmissions, the radio signals are treated as masks that obscure the actual noise distribution, and a deep network is used to estimate the noise distribution in the broadband radio frequency spectrum. This approach allows the SNR of the signals to be maximised in signal processing and, consequently, detection characterised by higher values of quality indicators.


This research was funded by the National Centre for Research and Development, grant number SZAFIR/05/A/027/01/2021 on “Development of data transmission technology and construction of a MF-TDMA modem for safe and reliable data transmission in VSAT communication systems” (acronym SatBit) as part of state defense and security program under competition no. 1/SZAFIR/2020.


Email Address of Submitting Author

ORCID of Submitting Author


Submitting Author's Institution

Military University of Technology in Warsaw

Submitting Author's Country

  • Poland