A Survey on AI-Driven Energy Optimization in Terrestrial Next Generation Radio Access Networks

This survey uncovers the tension between AI techniques designed for energy saving in
\nmobile networks and the energy demands those same techniques create. We compare modeling approaches
\nthat estimate power usage cost of current commercial terrestrial next-generation radio access network
\ndeployments. We then categorize emerging methods for reducing power usage by domain: time, frequency,
\npower, and spatial. Next, we conduct a timely review of studies that attempt to estimate the power usage
\nof the AI techniques themselves. We identify several gaps in the literature. Notably, real-world data for
\nthe power consumption is difficult to source due to commercial sensitivity. Comparing methods to reduce
\nenergy consumption is beyond challenging because of the diversity of system models and metrics. Crucially,
\nthe energy cost of AI techniques is often overlooked, though some studies provide estimates of algorithmic
\ncomplexity or run-time. We find that extracting even rough estimates of the operational energy cost of AI
\nmodels and data processing pipelines is complex. Overall, we find the current literature hinders a meaningful
\ncomparison between the energy savings from AI techniques and their associated energy costs. Finally,
\nwe discuss future research opportunities to uncover the utility of AI for energy saving.