Novel Methodological Proposal for Decision-Making in Decentralized Cognitive Radio Networks Based on Information Exchange Between SU

Cognitive Radio Networks (CRN) is a technology that avoids inefficient spectrum allocation and ensures efficient spectrum use. So, four processes are implemented: detection, analysis, decision-making, access, and adaptation. Despite the relevance of decision-making, it has not been explored to the same extent as the other processes. In CRNs, the decision-making process is developed according to the network architecture: centralized, distributed, and decentralized. Decentralized Cognitive Radio Networks (DCRN) are a hybrid model that uses the advantages of centralized and distributed networks simultaneously. Decentralized architectures have the infrastructure and are easy to implement. This decentralized approach is chiefly efficient for large networks and is considered the best option for public safety networks and social networking services. In order to address the challenges associated with DCRN decision-making and contribute to developing more effective approaches, this paper proposes a novel methodology for DCRN decision-making. A simulation environment for DCRN based on actual spectral occupancy data is developed, the performance of three Multi-Criteria Decision-Making Techniques (MCDM) in such a simulation environment is analyzed, and an information-sharing strategy between users is proposed. In order to assess the performance, two QoS metrics were used: the cumulative number of handoffs and the cumulative number of failed handoffs. The results obtained display a balance in which all users benefit. Although not all users get the maximum gain, all users contribute to reducing the number of channel changes and decreasing interference with other users.