Low-Carbon, User-Centric Networks Lead 5G Era

As global climate change accelerates and extreme weather events become more frequent, there is an expanding need across industries for sustainable, low-carbon growth. Digitization is essential to achieving carbon neutrality and is crucial to advancing global efforts to address climate change. With the widespread deployment of 5G base stations, managing energy usage goes beyond mere energy conservation—it involves striking a balance between service expansion and user experience. This means implementing energy-saving mechanisms that prioritize real user requirements and experiences, aiming to minimize energy consumption per bit while maintaining robust network performance.

HI-RAN: Enhancing Energy Efficiency Through Intelligent Coordination with a Distributed Architecture

Energy-saving technologies for mobile network have undergone extensive innovations in time-domain symbol shutdowns, frequency-domain carrier shutdowns, spatial-domain channel shutdowns, power-domain management, and even extreme dormancy in device domains. However, factors such as multi-mode, multi-frequency networks, increasingly complex deployment scenarios, and diverse service expectations make it challenging to implement precise and efficient energy-saving solutions that are site specific and time specific.

ZTE’s hybrid intelligent RAN (HI-RAN) energy-saving solution utilizes a dual-layer computational architecture that encompasses both the network management layer and the base station native layer. The distributed architecture allocates AI’s three fundamental aspects—data annotation, model training, and inference execution across its layers, creating a system where model training takes place at the network management layer, while inference is executed within the base station. The solution reduces energy consumption and enhances energy efficiency without compromising network performance, employing technologies such as predictive optimization, autonomous orchestration, multi-frequency collaboration and digital twin visualization.

The solution leverages the collection of network performance and energy consumption data to enable traffic load forecasting and implement optimal energy-saving strategies in response to variations in service or network performance:

Traffic load forecast analysis: Employs deep learning to predict traffic loads at the cell level, identifying low-load cells suitable for energy conservation.

Service offloading capability analysis: Evaluates real-time measurement reports to identify co-coverage cells for energy conservation and forecast their load patterns to identify the optimal offloading (coverage compensation) cells.

Energy conservation strategy self-configuration: Creates energy conservation strategies based on specific requirements, such as shutdown thresholds, time windows, duration, and energy-saving features.

Self-optimizing strategy iteration: Evaluates comprehensive traffic scenarios, energy conservation effects, and KPI trends to enhance self-learning and continuously update strategies online, ultimately achieving an optimal balance between energy usage and network performance.

To prevent sudden traffic surges or declines in user experience during energy-saving operations, base stations must rapidly detect network performance degradation and adjust energy-saving strategies in real time. Continuous evaluations are conducted across three phases—pre-, during, and post-energy conservation—to ensure predictable network performance indicators in energy-saving areas through a "real-time, event-driven, and periodic" performance assurance mechanism.

Pre-energy conservation: Monitors control and user plane indicators of both energy-saving cell and basic coverage cells. If established targets are not met, restrictions on access are not applied.

During energy conservation: Monitors control plane indicators during cell access restrictions and user relocation. If performance indicators exceed the predicted thresholds, the energy-saving mode is exited.

Post-energy conservation: Monitors the control and user plane indicators of the basic coverage cell immediately after the initiation of energy-saving strategies. If performance indicators exceed the predicted thresholds, the energy-saving mode is exited.

Enhancing Energy-Saving Solutions Based on User Experience

The rapid development of mobile networks has introduced a variety of services, such as video streaming, online gaming, social media, and online shopping. Mobile networks play a crucial role in ensuring stable and reliable internet access, while delivering high-quality experiences, which increases the demand on the base stations' capacity to identify and meet service requirements. To support AI applications in wireless networks, base stations need expanded processing and storage capacities, including the installation of integrated computing and communication boards, to meet the demands of AI modeling.

Enhancements in energy-saving technologies, based on service experience, utilize base station computing capabilities to collect service data, monitor the patterns of popular applications, and promptly adapt to the changes in these patterns. This ensures a higher-quality experience for essential services and advances application-level energy-saving technologies.

Latency-sensitive services: Latency-sensitive applications, such as real-time video calling, online gaming, and cloud meetings, require minimal network delay and consistent transmission of data. Even a small increase in latency may negatively impact the user experience, resulting in video freezes, audio interruptions, or gaming lags. In providing these services, base station systems adopt a minimal data joint scheduling delay threshold and, if necessary, may skip joint scheduling to ensure fast transmission of information to the user. This approach ensures seamless real-time service while keeping network stability.

 Latency-insensitive services: For services without real-time response requirements, such as file downloads, large data synchronization, and system upgrades, users do not need immediate feedback. This allows greater flexibility in energy-saving techniques. The network may adopt higher data joint scheduling delay thresholds, allowing a greater number of data packets to be aggregated before transmission, thus extending the symbol shutdown duration. This significantly decreases power amplifier energy usage and optimizes overall system energy efficiency.

Future mobile networks will face greater challenges due to growing demand for new services, such as holographic communication, intelligent interaction, and digital twins. The explosive growth in network traffic will raise energy consumption requirements. By integrating intelligent energy-saving technologies with user-centric improvements, 5G and future networks will substantially improve network efficiency while lowering overall energy consumption. This is going to propel the global digital society towards a low-carbon, sustainable future, offering strong support for achieving "mobile net-zero".