Connected map-induced resource allocation scheme for cognitive radio network quality of service maximization

Abstract

Abstract Quality of Service (QoS) in cognitive radio networks (CRNs) is achieved through fair resource allocation and scheduling for secondary users regardless of channel capacity through multi-channel communications. Fairness index updates are periodic towards multi-user allocations to meet the QoS demands. In this article, a Connected Resource Map-induced Resource Allocation Scheme (CRM-RAS) is introduced. The proposed scheme identifies radio and user resource availability and constructs an allocation map from the primary users. For a periodic allocation interval, the map’s fairness index is updated through maximum resource utilization and QoS factor. This QoS factor is computed based on low latency and high allocation rates that are directly proportional to the fairness index. The fairness index is verified using distributed federated learning that is active between the primary and secondary user terminals. If the fairness index drops below the actual allocation rate, then the scheduling for resource allocation with concurrency is pursued. Based on the improving fairness index through concurrent scheduling the distributed federated learning encourages consecutive radio resource allocation. Thus the process is repeated until the allocation map is confined to a one-to-one connectivity between the primary and secondary users. The proposed CRM-RAS achieves 8.15% high sum rate and 8.88% less error rate for the maximum SNR.

Keywords

• CRN
• Federated learning
• QoS
• Resource allocation
• Resource scheduling
• SDG 9: Industry, Innovation, and Infrastructure