6G cognitive radio: Optimizing resources with cluster-assisted downlink hybrid multiple access

The surge in educational, entertainment, multimedia, and gaming applications used by the number of users necessitates substantial data rates, compelling the exploration of the untapped spectrum in the forthcoming sixth-generation (6G) networks. Traditional networks fall short of meeting the escalating demands, prompting the integration of cognitive radio networks (CRNs) in 6G. Cognitive radio (CR) technology enhances spectrum utilization by opportunistically accessing unused spectrum when not in use by licensed users. In CRN, primary users (PUs) are served by the primary radio base station (PBS), while the secondary radio base station (SBS) utilizes opportunistic spectrum access using spectrum sensing, catering specifically to secondary users (SUs). Hybrid multiple access (HMA) as a combination of orthogonal multiple access (OMA) and non-orthogonal multiple access (NOMA), is introduced to further accommodate the increasing number of users in 6G, ultimately enhancing spectral efficiency. Notably, this paper investigates a novel CR based cluster assisted downlink HMA (CR-CDHMA) in CRN to jointly optimize the network rate, user admission in clusters, user association with SBS, user-channel allocation obtained through opportunistic spectrum access using spectrum sensing, and fairness in power allocation. A novel mathematical problem based on the proposed network model is formulated as a mixed-integer nonlinear programming (MINLP) problem, addressed through an $ϵ$-optimal outer approximation algorithm (OAA). Extensive simulations confirm the novel proposed technique’s effectiveness in 6G CRN rate improvement, user admission in clusters, user association with SBS, user-channel allocation obtained through opportunistic spectrum access using spectrum sensing, and fairness in power allocation under the constraints of false-alarm, missed-detection, and quality of service (QoS) surpassing performance compared to the existing OMA only assisted CRN and NOMA-only assisted CRN. An $ϵ$-optimal algorithm achieves results with $ϵ=10^{-3}$, demonstrating its computational efficiency.