Tuesday , August 14 2018

Improved Energy Detector for Spectrum Sensing Using
Neuro-Fuzzy Double Threshold Technique

Barnali DEY1, Ashraf HOSSAIN1, Valentina. E. BALAS2*, R. N. BERA3
1 Department of ECE, National Institute of Technology
Silchar, Assam, India
e-mail: ashraf@ece.nits.ac.in, barnali.mou.dey@gmail.com.
2 Department of Automation and Applied informatics, Aurel Vlaicu University of Arad,
310130, Romania
valentina.balas@uav.ro (*Corresponding author)
3 Department of ECE, Sikkim Manipal Institute of Technology, Sikkim, India
e-mail: rbera@smit.smu.edu.in

ABSTRACT: This paper reports a modified and computationally efficient energy detection method for spectrum sensing in cognitive radio (CR) system utilising hybrid soft computing paradigms. The proposed soft computing method uses artificial neural network and fuzzy logic for learning and decision making as a solution to the problems when CR system is subjected to threshold uncertainty and uncertainties in noise variance. The proposed soft computing method should be referred thereafter as Neuro-Fuzzy double threshold (NFDT) technique. The developed technique utilizes the concept of double sensing credibility while formulating the fuzzy decision rule and artificial neural network to learn the threshold function. Through simulation results it is validated that proposed method leads to overall improved probability of detection compared to conventional and existing methods.

KEYWORDS: Spectrum Sensing, Double threshold, Energy detector, Cognitive Radio (CR), Neuro-Fuzzy technique.

>>FULL TEXT: PDF

CITE THIS PAPER AS:
Barnali DEY, Ashraf HOSSAIN, Valentina. E. BALAS*, R. N. BERA,
Improved Energy Detector for Spectrum Sensing Using Neuro-Fuzzy Double Threshold Technique, Studies in Informatics and Control, ISSN 1220-1766, vol. 26(3), pp. 335-342, 2017.

https://doi.org/10.24846/v26i3y201709