Wi-Fi Meets ML: A Survey on Improving IEEE 802.11 Performance With Machine Learning
This survey adopts a structured approach to describe the various Wi-Fi areas where ML is applied and identifies specific open challenges and general future research directions, providing readers with an overview of the main trends.
Abstract
Wireless local area networks (WLANs) empowered by IEEE 802.11 (Wi-Fi) hold a\ndominant position in providing Internet access thanks to their freedom of\ndeployment and configuration as well as the existence of affordable and highly\ninteroperable devices. The Wi-Fi community is currently deploying Wi-Fi 6 and\ndeveloping Wi-Fi 7, which will bring higher data rates, better multi-user and\nmulti-AP support, and, most importantly, improved configuration flexibility.\nThese technical innovations, including the plethora of configuration\nparameters, are making next-generation WLANs exceedingly complex as the\ndependencies between parameters and their joint optimization usually have a\nnon-linear impact on network performance. The complexity is further increased\nin the case of dense deployments and coexistence in shared bands. While\nclassical optimization approaches fail in such conditions, machine learning\n(ML) is able to handle complexity. Much research has been published on using ML\nto improve Wi-Fi performance and solutions are slowly being adopted in existing\ndeployments. In this survey, we adopt a structured approach to describe the\nvarious Wi-Fi areas where ML is applied. To this end, we analyze over 250\npapers in the field, providing readers with an overview of the main trends.\nBased on this review, we identify specific open challenges and provide general\nfuture research directions.\n