Sixth Generation Enabling Technologies and Machine Learning Intersection: A Performance Optimization Perspective

Emmanuel Ekene Okere, Vipin Balyan

The fifth generation (5G) of wireless communication is in its finalization stage and has received favorable reception in many nations. However, research is now geared towards the anticipated sixth-generation (6G) wireless network. The new 6G promises even more severe performance criteria than the current 5G generation. New sophisticated technologies and paradigms are expected to be incorporated into the 6G network designs and procedures to meet the ever-dynamic user needs and standards. These 6G-enabling technologies include digital twin (DT), intelligent reflecting surface (IRS), visible light communication (VLC), quantum computing (QC), blockchain, unmanned aerial vehicles (UAVs), and non-orthogonal multiple access (NOMA), among others. Optimal network performance requires that machine learning (ML) techniques be integrated over the 6G wireless network to provide solutions to highly complex networking problems, massive users, high overhead, and computational complexity. Consequently, we have provided a state-of-the-art overview of wireless network generations leading to the future 6G, and huge emphases have been laid on ML’s role in optimization applications for different enabling 6G technologies. Several key performance indicators for the different application scenarios have been highlighted. ML has proved to significantly improve the performance of the existing 6G-enabling technologies, and choosing the appropriate approach can ultimately yield optimal results.