Radio resource allocation in a 6G D-OMA network with imperfect SIC: A framework aided by a bi-objective hyper-heuristic

• Published in: Engineering applications of artificial intelligence 2023-03, Vol.119, p.105830, Article 105830
• Main Authors: Torres, Fábio de O., Júnior, Valdivino A. Santiago, Costa, D.B. da, Cardoso, Diego L., Oliveira, Roberto C.L.
• Format: Article
• Language: English
• Subjects: Bi-objective selection hyper-heuristic; D-OMA; Imperfect SIC; QoE
• ISSN: 0952-1976; 1873-6769
• DOI: 10.1016/j.engappai.2023.105830

In the sixth generation (6G) of mobile communication networks, radio resource allocation management (RRAM) systems must offer high levels of customization and sustainability in their operations. For this reason, this article proposes a framework that maximizes the quality of experience (QoE) of users of a 6G network that implements the delta-orthogonal multiple access (D-OMA) scheme as well as the system’s energy efficiency (EE), considering imperfections in successive interference cancellation (SIC) activities. The framework has two phases: in phase (a), a low-complexity heuristic was implemented in order to carry out the selection of base stations (BSs) and distribution of the electromagnetic spectrum slices to each communication device (CD); and in phase (b), a new bi-objective selection hyper-heuristic (SHH) performs the transmission power levels allocation. To evaluate the framework’s performance, we performed two tests: First, we evaluated the generalization capacity of our SHH with the isolated application of its constituent metaheuristics. Second, our hyper-heuristic was compared to other state-of-the-art transmit power allocation methods (TPLAMs) developed for non-orthogonal multiple access (NOMA) environments. For a fair comparison, all TPLAMs used the same heuristic for phase (a). As a result, the framework, in addition to providing high levels of QoE to network users, also presented low levels of failure probability as well as improved the average EE of the system.