Next Generation Multiple Access Techniques for Wireless Communications

The wireless networks evolution, is marked by diverse and complex use cases. Key categories like massive machine-type communications (mMTC) and ultra-reliable low-latency communications (URLLC) present unique challenges in the internet-of-things (IoT) ecosystem. Concurrently, enhanced mobile broadband (eMBB) applications demand high data rates, especially in densely populated areas. These varying requirements underscore the need for specialized multiple access schemes tailored to each use case. The thesis explores various approaches to address these challenges. Firstly, two uplink orthogonal multiple access (OMA) solutions are discussed. A semi-grant-free scheme wherein the scheduling is optimized based on the partial knowledge of the user activation is proposed. Then, the study extends to correlated packet generations, developing a distributed learning-based scheduler, which outperforms existing state-of-the-art correlation-based schemes in scenarios with moderate traffic correlation and intensity. Secondly, for downlink OMA, the thesis considers a millimeter wave (mmWave) scenario wherein an intelligent reflecting surface (IRS) with practical (re)configuration constraints is used to mitigate the strong channel attenuation. Heuristic scheduling techniques for downlink time-division multiple-access (TDMA) and orthogonal frequency-division multiple access (OFDMA) are proposed and their performance is deeply discussed. Finally, the last part of the thesis focuses on non-orthogonal multiple access (NOMA) within the context of unsourced random access (URA). Novel tensor-based approaches for random access are presented, showing significant improvements in energy efficiency and fading robustness compared to state-of-the-art URA solutions.