Artificial neural network based spectrum recognition in cognitive radio

Cognitive radio is the best solution for spectrum scarcity and spectrum underutilization over wireless communication challenges. It empowers secondary users to use primary user's spectrum without any interference. In this paper, a new approach of artificial neural network based spectrum sensing is proposed which senses the availability of a vacant channel in the primary user's spectrum and allocates it to the secondary users. Here neural network is used because it is a smart intelligent phenomenon with the capabilities to understand the environment, learn and adjust in real time operating parameter according to specific need of unlicensed user. The primary objective of this paper is to maximize the decision accuracy under different noisy conditions and its implementation through the perceptron based neural network system that tremendously enhances the desirable throughput. Finally the throughput is evaluated on simulation basis and results shown using MATLAB.