Channel Activity Analysis of Cognitive Radio with PCA Preprocessing and Different Clustering Methods

Abstract

This article studies methods for channel activity analysis in cognitive radio that do not require a priori information about the signal, the channel and the noise power. Channel identification is a great challenge for cognitive devices, so channel activity plays an important role in spectrum management decisions. Dimension reduction with Principal Component Analysis (PCA) and various clustering methods (agglomerative, K-means and K-medoids) are used to reduce the hardware and software requirements while maintaining and improving detection accuracy. The goal of the proposed analysis is to discover the optimal parameters for data processing in spectrum hole allocation for cognitive radio systems by using preprocessing and cluster analysis.