L.1390 Energy saving technologies and best practices for 5G radio access network (RAN) equipment

Scope

L.1390 identifies energy-saving potentials, describes energy-saving principles and technologies for 5G RAN and related equipment, and provides best practice recommendations on their usage and control. It aims to reduce 5G RAN energy consumption, lower operational costs, and establish a green and high-efficiency network. Additionally, it proposes optimising and controlling these energy-saving technologies using AI, including an AI-driven overall architecture.

Summary

The BTS total power consumption at the site (P) is modelled by a first-order linear equation composed of a baseline power consumption (b) and variable power consumption (a × T) positively related to the load.

P = a × T + b

Hardware capabilities must be improved to reduce the proportion of baseline power consumption and enhance the equipment’s variable power consumption capability.

L.1390 categorises 5G energy-saving technologies into four domains:

  1. Time-domain energy saving (e.g., discontinuous transmission (DTX) deactivates PA in symbols where no information is transmitted),
  2. Spatial-domain energy saving (e.g., MIMO muting or RF channel shutdown),
  3. Frequency-domain energy saving (Large-scale: carrier shutdown; Small-scale: subcarrier shutdown) and
  4. Power-domain energy saving (reducing or scaling down the PA output power)
AI-based Energy Saving Architecture

L.1390 includes the AI overall architecture (see figure below) for a more precise energy-saving strategy based on site-specific traffic, Scene identification, Traffic forecast, Co-coverage identification, online iteration and parameter optimisation and other site-related conditions.

The AI algorithms use data from the network management layer. The outputs from the AI algorithms (aimed at balancing system performance and energy-saving effect) are fed back to the network management layer and used for configuring the BSs in the network element layer. Identifying service types and their energy efficiency differences allows real-time evaluation of service requirements to provide the service with a network of higher energy efficiency for overall optimisation.

Index