Treffer: A PYTHON-BASED FRAMEWORK FOR ADVANCED RESEARCH AND DEVELOPMENT ON SPECTRUM SENSING FOR COGNITIVE RADIO.

The work of this paper is addressing the issue of narrowband spectrum sensing (SS) applications for cognitive radio networks assuming the OFDM signal over noisy and various fading channels with focus on the development of an open-source simulation and development software. The paper also proposes a residual and recurrent convolutional Neural Network (NN)- based method for spectrum sensing. This method is compared with a baseline energy detection (ED) approach. Detection performance of the models is analysed given various Signal-to-Noise Ratio (SNR) values using three-base scenarios. First two scenarios correspond to an additive white gaussian noise (AWGN) channel respectively Rayleigh flat fading channel and the third one corresponds to the frequency-selective Rayleigh fading channel since the SNR value and fading can affect drastically the detection performance in terms of false positives/false negatives leading to erroneous estimators. The experimental results are analysed through the receiver operating characteristics (ROC) plot containing the curves for both ED models enhanced by the denoising methods. On average, the classic ED algorithm with dynamic threshold outperforms the NN-based model, especially in low SNR domains. The NN-based model trained on constrained, tailored dataset characteristics outperforms the classic ED model in scenarios described by the corresponding characteristics. [ABSTRACT FROM AUTHOR]

Copyright of Acta Technica Napocensis. Electronica-Telecomunicatii is the property of Universitatea Tehnica din Cluj-Napoca, Faculty of Electronics, Telecom & IT and its content may not be copied or emailed to multiple sites without the copyright holder's express written permission. Additionally, content may not be used with any artificial intelligence tools or machine learning technologies. However, users may print, download, or email articles for individual use. This abstract may be abridged. No warranty is given about the accuracy of the copy. Users should refer to the original published version of the material for the full abstract. (Copyright applies to all Abstracts.)