The two primary power delivery challenges with 5G new radio (NR) are improving operational efficiency and maximizing sleep time. For example, Ericsson estimates that 94% of the company’s carbon emissions are from the operation of its products. It’s been estimated that base station resources are generally unused 75 – 90% of the time, even on high-load networks.
The base station power consumption constituents are evolving, making the power challenges a moving target, as illustrated in Figure 1. For LTE networks, the power amplifiers (PA) were large power consumers, and the focus was on improving PA idle modes and energy efficiency.
That has been largely accomplished, and today’s 5G NR installations benefit from using microsleep modes where PAs are deactivated and reactivated in microseconds and energy-proportional computing. However, the growing use of massive MIMO (multiple-input, multiple-output) and increasing demand for high bandwidth drives PA (analog) power backup, resulting in an increased focus on efficient thermal management.

As those issues are addressed, mature 5G NR deployments are expected to increase the focus on lower carbon emissions and more compact designs. The evolving requirements of 5G NR power are being addressed in several ways, including:
- Increased use of renewable energy sources.
- Improved load dependence of system elements by continuous improvements in sleep modes, reducing reactivation times, and increasing granularity.
- Expanded use of advanced power management techniques like dynamic voltage and frequency scaling (DVFS).
- Artificial intelligence (AI) can be used to increase the impact of improved load dependence and power management technologies.
Renewable energy and 5G
There are two sides to the coin regarding renewable energy and 5G. Of course, 5G networks will be major consumers of renewable energy to reduce their carbon footprint. Solar panels or other renewable energy sources can directly power small cell 5G base stations.
In addition, 5G’s high bandwidth and low latency can enable real-time data collection from renewable energy farms, enabling better grid management and integration. 5G is also used in microgrids to facilitate more efficient coordination of green energy resources, storage facilities, and energy consumption.
Advanced sleep modes
4G has limited ability to selectively shut down devices without traffic. The radio interface in a 4G cell was required to transmit about 1,000 reference signals per second, even without an active mobile device connected. The “lean carrier” design in 5G supports more flexible signaling intervals and transmission-free time slots during low-traffic periods.
Advanced sleep modes (ASM) were added with 5G. ASM includes four sleep mode levels with different durations. The longer the duration, the more components are shut down. Durations include:
- 71 microseconds in SM1
- 1 millisecond in SM2
- 10 milliseconds in SM4
- 1 second in SM4 (this is standby mode where the base station is inactive, but the backhaul remains on)
DVFS
DFVS is a well-established power management technique that can be applied to PAs and digital devices like microcontrollers. However, it’s not as widely used as it could be due to implementation cost and complexity. That is changing in response to the need for more energy efficiency.
DVFS adjusts the operating voltage and clock frequency in real-time based on the current workload to adaptively balance performance and power consumption (Figure 2).

Artificial intelligence
Applications based on long short-term memory (LSTM), a recurrent neural network (RNN) type, are used in 5G networks to analyze and predict network traffic patterns, enabling dynamic allocation of resources, including power, for improved efficiency.
LSTM models are also being used to implement dynamic power control strategies. In these strategies, the transmit power of a cell tower is adjusted based on real-time traffic conditions, further optimizing energy usage. LSTM also identifies or predicts which cell towers will be underutilized, enabling the network to optimize their loading and increase cumulative energy savings.
Summary
Optimizing the sustainability of 5G NR operations is complex. Several tools, including renewable energy, advanced sleep modes, DVFS, and AI, are being deployed to improve sustainability.
References
5G base stations and the challenge of thermal management, Essentra Components
5G energy efficiency: A sustainable connectivity solution, AT&T Business
5G Energy Efficiency Overview, European Scientific Journal
A Long Short-Term Memory Network-Based Radio Resource Management for 5G Network, MDPI future internet
Efficient Approaches to 5G Power Challenges in the Telecom Industry, Vetiv
Energy Efficiency Concerns and Trends in Future 5G Network Infrastructures, MDPI energies
Improving energy performance in 5G networks and beyond, Ericsson
Power Amplifier Modules and Their Role in 5G Design, Qorvo
System level analysis of fast, per-core DVFS using on-chip switching regulators, ResearchGate
The Environmental Impact of 5G: Balancing Technological Advancements and Sustainability, Verizon
Three-Level Hybrid Envelope Tracking Supply Modulator with High-Bandwidth Wide-Output-Swing, MDPI applied sciences
WTWH related links
How to calculate RF power amplifier efficiency
Release 16: Enhancements in power savings can make your day
6G promises to bring sustainability to telecom
Optimize RF signal quality in 5G power amps
Can dielectric waveguide antennas boost 5G efficiency?