IMPACT OF THE THRESHOLD VALUE ON DETECTION PERFORMANCE IN EIGENVALUE BASED SENSING METHODS

Fatih Yavuz ILGIN¹ , Cebrail ÇİFTLİKLݲ

In today's communication systems, especially in wireless networks, have been proven as a result of measurements made in the radio frequency spectrum. This density means that there is space shortage for new services or technologies in the radio frequency spectrum. The main reason for this congestion in the frequency spectrum is the inefficient use of the radio frequency spectrum due to the fixed frequency assignment policies. Cognitive radio systems are the general name of technologies that have been developed in recent years to remedy the increasing spectral density. The purpose of this technology is to continuously monitor the radio frequency spectrum to evaluate free space. It is the initial phase of these technologies to perceive the frequency spectrum in the most accurate way in cognitive radio systems. Despite the use of very different methods in the literature for spectrum detection, eigenvalue spectrum detection is one of the most studied topics on the basis of some features. The selection of the threshold value directly affects the algorithm performance in the eigenvalue detection. In this study, the effect of different threshold values used for different eigenvalue detection methods on the algorithm performance is examined.

Keywords
Detection Theory, Cognitive Radio, Tracy-Widom Distribution, Covariance Matrix, Spectrum Efficiency,