Open AccessCooperative Bayesian‐based detection framework for spectrum sensing in cognitive radio networks
Author(s) -
Maleki Alireza,
Mirzahosseini Davood,
Moghaddam Naser Ahmadi
Publication year - 2019
Publication title -
iet communications
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.355
H-Index - 62
eISSN - 1751-8636
pISSN - 1751-8628
DOI - 10.1049/iet-com.2019.0353
Subject(s) - cognitive radio , computer science , variance (accounting) , noise (video) , bayesian probability , matlab , energy (signal processing) , spectrum (functional analysis) , algorithm , telecommunications , machine learning , artificial intelligence , data mining , statistics , wireless , mathematics , physics , image (mathematics) , accounting , quantum mechanics , business , operating system
In this study, a cognitive radio network is considered in which multiple secondary users intend to detect a primary user frequency band in order to specify whether it is occupied or not. To this end, a blind Bayesian framework is proposed by which secondary users cooperatively perform spectrum sensing. In practice, it is impossible to estimate the noise variance accurately (noise uncertainty problem) and this can degrade the performance of some previous spectrum sensing algorithms like energy detection (ER). To overcome this issue, unlike the conventional ER, the proposed algorithm utilises marginalisation to eliminate the effect of uncertainty in noise variance estimation. By computer simulations using MATLAB, it can be seen that the authors' algorithm reaches the ideal case for P faby improving the level of cooperation (increasing the number of secondary users) and yet its P mdis also improved compared to ER in practical situations (presence of noise uncertainty).