Robust Clustering of Ad-Hoc Cognitive Radio Networks under Opportunistic Spectrum Access

The time and space varying nature of channel availability among cognitive radio nodes challenges connectivity and robustness of ad-hoc cognitive radio networks. Clustering of neighbouring cognitive radio nodes is a suitable approach to address this challenge. A cluster utilizes the same channel for payload communication among the nodes. As a consequence, clustering enables cooperative spectrum sensing, supports a coordinated channel switching and simplifies routing in ad-hoc cognitive radio networks. However, the sudden appearance of primary nodes can lead to the loss of connectivity within a cluster or between clusters. This impact can be mitigated to some extent by the way clusters are formed. In this work we discuss a distributed, low-complexity clustering algorithm that emphasizes the robustness of clusters by improving inter- and intra-cluster connectivity. The algorithm is proven to converge fast while numerical evaluation shows a significant improvement of robustness compared to related work.