NON-LINEAR SPECTRUM SENSING SCHEME IN IMPULSIVE NOISE ENVIRONMENTS FOR COGNITIVE RADIO

Abstract

Recently, prior studies have addressed and solved the issues of spectrum sensing through Cognitive
Radio (CR) technology, assuming that the noise is Gaussian. However, in real-world scenarios, this assumption
often falls short as various types of noise exhibit non-Gaussian and impulsive characteristics. Consequently,
there is significant value in investigating spectrum sensing techniques that are specifically tailored to operate
effectively in impulsive noise presence. Hence, within the context of CR networks operating in impulsive noise
environments, multiple reception antennas are used to tackle the challenge of spectrum sensing. To efficiently
deal with the unique characteristics of impulsive noise, which manifest as a heavy-tailed probability density
function, an innovative non-linear combining scheme is introduced, founded on the arranging statistics. The
simulation results confirm that the introduced scheme consistently offers robust performance and stability in
spectrum sensing and outperforms conventional methods in terms of detection performance, up to 65%, within
impulsive noise environments.