A channel availability classification for Cognitive Radio Networks using a monitoring network

In Cognitive Radio (CR) networks, the accurate information gathering for the channel availability brings some major challenges. It is seen that the availability should be classified effectively and opportunistically considering the channel utilization levels of the licensed users. Moreover, local spectrum monitoring mechanisms should also be used for accurate determination of the channel availability. Considering these challenges, in this paper, an estimation and correlation based scheme is proposed to build the availability map of the licensed channels according to the utilization requirements of the CR users. The proposed scheme is aided by a monitoring network for more accurate and robust detection of the available spectrum. The monitoring network is only responsible of sensing the channels and estimating the availability using auto-regression analysis. The monitoring nodes disseminate the estimations to the centralized entity, who builds the availability map for the channels to be used by the CR users. This classification map is prepared by considering the characteristics of PU channels and it is based on the cross-correlation analysis of the estimations by the monitoring nodes. The evaluations show that the proposed estimation and correlation based scheme improves the estimation accuracy of the licensed channels as well as the overall utilization performance compared to the fully-cooperative traditional scheme.