Abstract
Energy efficiency (EE) metrics are important tools to support evaluation and management of communication networks, and are of key interest in the development of the upcoming 6G network strategy. Because of their utility, EE metrics are widely used by network operators, in standards and in research. However, metrics suit varying evaluation purposes more or less well, yet are not always presented or applied consistently; for example when predicting future EC based on previously measured energy values.
With this in mind, we provide a classification of existing EE metrics and how they differ; including energy intensity (EI), bitper-joule efficiency, consumption-related EE, and output-related EE. We illustrate their use and limitations through the micro view of an idealized 6G base station (BS). Additionally, we also consider the application of EE metrics to evaluate the macro view of a number of BSs providing coverage in a certain area. In this case, EE metrics are used as a tool to evaluate systemlevel properties. Specifically, we illustrate their use to evaluate how BS sleep control can approximate an energy-proportional ideal system in a high-load regime.
With this in mind, we provide a classification of existing EE metrics and how they differ; including energy intensity (EI), bitper-joule efficiency, consumption-related EE, and output-related EE. We illustrate their use and limitations through the micro view of an idealized 6G base station (BS). Additionally, we also consider the application of EE metrics to evaluate the macro view of a number of BSs providing coverage in a certain area. In this case, EE metrics are used as a tool to evaluate systemlevel properties. Specifically, we illustrate their use to evaluate how BS sleep control can approximate an energy-proportional ideal system in a high-load regime.
Original language | English |
---|---|
Journal | IEEE Access |
Publication status | Accepted/In press - 27 May 2025 |
Research Groups and Themes
- ESRC Centre for Sociodigital Futures