Desempenho do sensoriamento espectral com fusão eficiente de decisões sob ruído impulsivo e imprecisão dos parâmetros estimados

Abstract

The cooperative spectral sensing (CSS) in cognitive radios (CR) networks presents a good alternative to mitigate imperfections at local sensing which are caused by thermal noise, fading, shadowing and propagation loss. In the CSS system, sensing information is transmitted to the fusion center (FC) by multiple cognitive radios in order to achieve a better performance at the decision making process regarding the state of the sensed channel. However, the CSS system increases the spectral resources requirements given that the sensing information is transmitted to the FC on a common channel. Recently, a novel decision fusion rule was proposed to increase spectral efficiency of CSS networks. In this rule, CRs are allowed report their local decisions to the FC at the same time-slot and with the same carrier frequency. The performance of this rule may be affected by FC’s imperfect knowledge of some system parameters, such as the common channel gains of each CR and the thermal noise variance at the receiver. This work aims to evaluate the performance of the decision fusion rule with imperfect channel gains and thermal noise variance estimates. Moreover, we analyze the performance of this rule when the noise has impulsive components, very common in real communications systems. All results are generated by simulations with a practical receiver implementation on the software VisSim/CommTM.