A Low-Interference Channel Status Prediction Algorithm for Instantaneous Spectrum Access in Cognitive Radio Networks

Abstract

Various channel status prediction algorithms in cognitive radio networks have been proposed to predict channel availability. However, the existing algorithms predict imperfectly and may cause interference to incumbent users, especially when the prediction output is immediately used to access the channels as it is in spectrum mobility. In this paper, we analyze the prediction errors of the conventional channel status prediction algorithm and propose a novel low-interference channel status prediction algorithm (LICSPA) that suppresses the interference caused by prediction errors. Simulation results show that the proposed LICSPA can effectively decrease interference while keeping the prediction accuracy high.