Estıimation of Gauss Mixing Models with Expectation Maximum in Cognitive Radios

Nowadays, it is known that there is a constantly increasing demand for the radio frequency spectrum. The biggest reason for this situation is the constantly increasing data sizes in wireless communication systems. Therefore, Cognitive Radio (BR) systems are potential technologies that can be a solution to future spectrum shortage problems. In the BR systems, the probability distribution function of the signal to be detected is utilized at the basis of spectrum detection. The probability distribution functions of the signals perceived by BR users according to their spatial location, which will perform spectrum detection in wireless communication systems, vary. The purpose of this study is to predict the common Gaussian mix model of signals perceived by different BR users using the expectation maximization algorithm. In the study, simulation results were performed according to different noise levels and different number of cognitive radio users in terms of spatial differences of different BR users. Thus, the estimated Gaussian mix models can be used to predict different spectrum sensing models.