Channel allocation in Cognitive Radio networks based on a predictive and adaptive database

According to IEEE 802.22, a cognitive base station uses the spectrum manager (SM) entity to control all the secondary users (SUs) and channels under its coverage area. This work proposes a novel approach for channel allocation in Cognitive Radio (CR) networks, through the use of a predictive and adaptive database. The aim of the proposed method is to improve the performance of the SM on the channel allocation procedure in cognitive radio networks. The SM uses the predicted data, obtained through a spectrum-modeling algorithm that emulates channel utilization, in order to allocate those channels with higher probability of being idle. The adaptive characteristic is ensured through periodic updates of the database, which are based on channel sensing results. Thus, channels are ranked so that those with best performance are preferred for SU allocation. Simulation results show that the use of the predictive and adaptive features presents better results than methods that do not consider such features. Furthermore, the results show that the number of channel sensing operations performed by the cognitive devices is greatly reduced when using the proposed channel allocation method, which improves the overall network performance.