Coalition Formation for Outsourced Spectrum Sensing in Cognitive Radio Network

Abstract

We inspect an opportunistic Cognitive Radio Network formed between a Primary Network Provider (PNP) and secondary IoT operators. In our system, spectrum-sensing duty is offloaded to a Sensing Agent (SA), which allows the IoT operators to improve data rate and reduce energy consumption otherwise needed for spectrum-sensing. IoT operators pay PNP for accessing licensed spectrum and SA for providing sensing as a service. We use Nash bargaining to finalize negotiations among PNP, IoT operators, and SA. Our investigation shows that incorporating SA improves the throughput and utility of PNP and IoT operators. Coalition formation among IoT operators helps them avoid intra-coalitional interference by accessing the spectrum in time-shared fashion. As the valuation of a coalition depends on other coalitions, we analyze coalition formation as a Partition Function Game. We emphasize economic fairness among the coalitions to prevent monopolization. We define a proportional fairness-based metric based on which the final coalition structure is determined. In general, finding an optimal coalition structure is an NP-hard problem. The fairness criterion, along with the unique properties of the Nash bargaining solution allows us to transform the coalition formation problem into a more tractable Number Partitioning Problem and develop an efficiently computable coalition formation technique.