Robust clustering for ad hoc cognitive radio network. Issue 5 (21st February 2018)

• Record Type: Journal Article
• Authors: Li, Di; Fang, Erwin; Gross, James
• Is Part Of: Transactions on emerging telecommunications technologies. Volume 29:Issue 5(2018)
• Page Start: n/a
• Page End: n/a
• Publication Date: 2018-02-21
• Subjects: Telecommunication -- Periodicals; 384.05
• DOI: 10.1002/ett.3285
• Languages: English
• ISSNs: 2161-5748

Abstract

Abstract: Cluster structure in cognitive radio (CR) networks facilitates cooperative spectrum sensing, routing, and other functionalities. Unlicensed channels, which are temporally available for a group of CR users in one area, consolidate the group into a cluster. More available unlicensed channels in a cluster make the cluster more likely to uphold against the licensed users' influence, making clusters more robust. This paper analyzes the problem of how to form robust clusters in a CR network such that CR systems benefit from collaboration within clusters despite intense primary user activity. We give a formal description of the robust clustering problem, prove it to be NP‐hard, and propose both centralized and distributed solutions. The congestion game model is adopted to analyze the process of cluster formation, which not only contributes to the design of the distributed clustering scheme but also provides a guarantee on the convergence to a Nash equilibrium and the convergence speed. The proposed distributed clustering scheme outperforms state‐of‐the‐art in terms of cluster robustness, convergence speed, and overhead. Extensive simulations are presented supporting the theoretical claims. Abstract : This paper analyses the problem of how to form robust clusters in a cognitive radio network such that cognitive radio systems benefit from collaboration within clusters despite intense primary user activity. The robust clustering problem is proved to be NP‐hard and both centralized and distributed solutions are proposed. The proposed distributed clustering scheme outperforms state‐of‐the‐art related works in terms of cluster robustness, convergence speed and overhead.