Feature Selection Based Spectrum Decision Making For Cognitive Radio Networks

Abstract

The growing success of new wireless  applications and services has led to  overcrowded  licensed  bands. Cognitive radio networks  have  been  proposed  as  a  solution  to  both  spectrum inefficiency  and  spectrum   scarcity  problems  by  improving  the utilization of limited radio resources. Cognitive radio networks are intelligent networks that can automatically sense the environment and adapt the communication parameters accordingly. These types of  networks  have  applications  in  dynamic  spectrum  access,  co- existence of different wireless networks, interference management etc. A spectrum decision framework is proposed to determine a set of spectrum  bands by considering the application requirements as well as the dynamic nature of the spectrum bands. The use of cyclic feature   based   methods   for   distributed   signal   detection   and classification   is   discussed   and   recent   results   are   presented. Moreover, a dynamic resource management scheme is developed to coordinate the spectrum decision adaptively dependent on the time- varying cognitive radio network capacity. Simulation results can be obtained  by  using  NS2  which  shows  that  the  proposed  method provide  efficient  bandwidth  utilization  while  satisfying   service requirements.