Joint opportunistic MIMO-mode selection and channel–user assignment for improved throughput in beyond 5G networks.

Cognitive radio (CR) technology enables dynamic spectrum access for 5G and Beyond 5G (B5G) networks. Specifically, CR provides massive spectrum opportunities for enabling large-scale deployment of 5G/B5G networks. Multiple-input, multiple-output (MIMO) technology is another promising technique that can significantly improve the performance of 5G/B5G networks. Integrating the MIMO technology with the CR technology can further improve the wireless network's performance. However, due to the dynamic nature of the CR operating environment, such integration imposes several design challenges regarding the appropriate MIMO-mode selection, spectrum assignment, and user selection. Unlike the majority of previously developed MIMO-based spectrum access protocols that execute the channel assignment sequentially, we investigate the problem of maximizing the network throughput by simultaneously providing distributed CR-aware MIMO-mode channel-assignment decisions for multiple CR transmissions (batching approach) subject to a set of constraints. The joint batch-based MIMO-mode and channel-assignment optimization is, in general, an NP-hard combinatorial integer non-linear problem. To solve such a joint problem in polynomial-time, we propose a two-stage optimization algorithm. In the first stage, the proposed algorithm finds the highest number of packets that each contending CR user can send over each idle channel. Then, in the second stage, the best channel–user assignment that can result in the highest number of transmitted packets is computed. To realize the proposed batch-based algorithm in a distributed manner, we adopt an admission control process that allows CR users to announce the required control information. Compared with the existing MIMO-based CRN protocols, the results show that by jointly performing batch-based channel assignment and per-user MIMO-mode optimization, the proposed protocol significantly enhances the overall network throughput by up to 60%. [ABSTRACT FROM AUTHOR]