Efecto de la distribución del tiempo de servicio de los usuarios primarios en las estadísticas de los espacios en blanco de un sistema de radio cognoscitivo

Abstract

Cognitive radio technology was developed for spectral efficiency improvement in mobile communication networks. This is achieved by allowing secondary users to opportunistically and transparently use the white spaces of the primary network. In this paper, the effect of the service time distribution of the primary users in the statistics of the white spaces is analyzed. In particular, the first two standardized moments of both, white space time and white space interarrival time are found, considering the following distributions for the primary service time: log-normal, Weibull, and Pareto. One of the most relevant results is that, for low (moderate or high) traffic load, the exponential (hiper-exponential) distribution is an excellent option for modelling the white space duration. Characterizing these variables allow us to use the on-off paradigm to capture the primary channel activity and, in this way, evaluate the performance of cognitive radio networks. Results were obtained by using discrete event simulation techniques.