Application of Heuristic-Learning Model to Reduce Spectrum Sensing Energy in Cognitive Radio Network

Abstract

Cognitive radio (CR) can access licensed spectrum opportunistically without creating any interference to the licensed users. This is possible due to frequent spectrum sensing to identify the underutilized spectrum bands. However, the behavior of spectrum sensing consumes remarkable amount of battery power and thus reduces the lifetime of a user. Though the primary concept of CR is to enhance spectrum utilization, the importance of energy efficiency brings several new challenges. For a user with limited battery power, better throughout and energy efficiency can be paradoxical. In this work, a low complexity heuristic approach is proposed along with a prediction method based on learning. This approach reduces energy consumption by avoiding unnecessary sensing processes according to the prediction. The significances of our proposed approach are shown through simulations.