Threshold Optimization in Maximum–Minimum Eigenvalue-Based Detection in Cognitive Radio Using Ant Colony Optimization

Abstract

Today's communication systems are concentrated majorly in wireless networks that depend upon the density in the radio frequency (RF) spectrum. This density means a shortage of space for new services or technologies to be created in the RF spectrum. The main reason for this density in the frequency spectrum is the inefficient use of the RF spectrum due to fixed frequency assignment policies. Cognitive radio systems, in contrast, are the general name of technologies developed to find a solution to this increasing spectrum density in recent years. The purpose of this technology is to evaluate the empty spaces by constantly monitoring the RF spectrum. Perceiving the frequency spectrum in the most accurate way in cognitive radio systems is the initial stage of these technologies. Although many different methods are used in the literature for spectrum detection, eigenvalues-based spectrum detection is among the most studied topic due to its features. The choice of threshold value in eigenvalues-based detection directly affects algorithm performance. In this study, the effect of threshold values used for eigenvalues-based detection method on algorithm performance was investigated using Ant Colony Optimization. Simulation results on the basis of probability of detection and SNR performance validate the importance of this research.