Spectral Gap Filling in Cognitive Networks: A Cooperative Game-Theoretic Approach

An optimal joint channel selection and power control scheme is investigated in a cognitive network context, where the cognitive network is composed by multiple cognitive interference channels. Here, we take the fairness among multiple secondary users (SUs) and Pareto optimality measured by the capacity maximization into consideration. The complex cooperation and competition relationship among multiple SUs and primary users (PUs) is described with the refined signal-to-interference plus noise (SINR) definition. According to the Nash axioms from the Nash bargaining cooperative game, the newly built utility function is formulated, and the spectral gap-filling problem is formulated as cognitive capacity Nash product maximization (CCNPM). To improve the centralized algorithm design in in the cooperative game theory framework, we employ the dual decomposition technique to achieve the distributed bargaining approaches. The proposed approaches are with low implementation complexities and the little information exchange.