NG-RAN Energy Cost Mapping for Coordinated AI/ML Energy Saving

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Existing AI/ML-based network energy saving systems face challenges in optimizing overall energy efficiency across multiple gNBs due to the lack of a standardized mapping rule from energy consumption to energy cost index, leading to ambiguities and suboptimal energy management decisions.

Innovation Solution

Introduce an AI/ML-driven energy cost index metric that is exchanged between NG-RAN nodes, using a unified mapping rule configured by the operator to normalize energy consumption values, allowing gNBs to make informed energy-saving decisions based on historical patterns and traffic offloading.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Adaptability or versatility

If AI/ML-based energy saving techniques are implemented without a standardized mapping rule, then energy management flexibility is improved, but measurement precision and decision accuracy deteriorate due to ambiguities in energy cost interpretation

Engineering Contradiction:

Improveenergy management flexibilityVSAvoidenergy cost interpretation accuracy

Core Design Contradiction:

Adaptability or versatilityVSMeasurement precision

Solution Approach 1:

The patent transforms raw energy consumption measurements into standardized energy cost indices through configurable mapping rules. This parameter transformation enables consistent interpretation of energy costs across different gNBs while maintaining the flexibility to adjust mapping configurations based on operator policies and network conditions.

Inventive Principle:

Principle #35Parameter changes

Solution Approach 2:

The patent creates an equipotential framework by establishing a common energy cost index baseline across multiple gNBs. Through standardized mapping rules, all gNBs operate from the same reference point for energy cost interpretation, eliminating ambiguities and enabling accurate comparative analysis for load balancing and traffic offloading decisions.

Inventive Principle:

Principle #12Equipotentiality

2Area of stationary object

If energy consumption measurements are collected from multiple gNBs without unified mapping rules, then data collection coverage is improved, but manufacturing precision and consistency deteriorate due to varying energy cost interpretations

Engineering Contradiction:

Improvedata collection coverageVSAvoidenergy cost consistency

Core Design Contradiction:

Area of stationary objectVSManufacturing precision

Solution Approach 1:

The patent implements a universal mapping rule framework that can be applied across all gNBs in the network. The standardized energy cost index methodology serves multiple functions simultaneously: it enables consistent energy consumption comparison, supports load balancing decisions, facilitates traffic offloading optimization, and provides a common basis for energy efficiency KPI calculation across diverse network scenarios.

Inventive Principle:

Principle #6Universality (Multi-functionality)

Solution Approach 2:

The patent applies parameter transformation by converting heterogeneous energy consumption measurements from different gNBs into a unified energy cost index scale. This standardized parameter representation ensures consistent interpretation across the entire network while preserving the ability to collect and analyze energy data from multiple sources.

Inventive Principle:

Principle #35Parameter changes

3Speed

If gNBs make energy saving decisions based on unnormalized energy consumption values, then decision speed is improved, but productivity and overall energy efficiency deteriorate due to suboptimal energy management

Engineering Contradiction:

Improvedecision speedVSAvoidoverall energy efficiency

Core Design Contradiction:

SpeedVSProductivity

Solution Approach 1:

The patent implements preliminary normalization of energy consumption data into standardized energy cost indices before energy saving decisions are made. By pre-processing and standardizing energy cost information across all gNBs, the system enables rapid comparison and decision-making while ensuring that decisions are based on accurate, normalized data that reflects true energy efficiency differences.

Inventive Principle:

Principle #10Preliminary action

Solution Approach 2:

The patent establishes a feedback mechanism where energy cost indices are continuously measured, standardized through mapping rules, and used to inform energy saving decisions. The results of these decisions are fed back into the system, allowing for continuous optimization of energy efficiency while maintaining fast decision-making through the use of pre-established mapping relationships.

Inventive Principle:

Principle #23Feedback

Data Source

PatentUS20250212116A1Ai/ML-based network energy saving system

Publication Date: 2025.06.26 INTEL CORP

AI summary

Systems and methods are disclosed for a comprehensive framework for managing energy consumption to energy cost index mapping rules in next-generation radio access networks (NG-RAN) enables operators to configure unified mapping rules for groups of gNBs to optimize network energy efficiency through artificial intelligence/machine language (AI/ML)-based decisions. The framework includes multiple mapping approaches for both current and future releases, including linear mapping, mapping tables, and enhanced mapping methods that incorporate additional load parameters. The management service architecture supports both standalone and embedded management functions, allowing operators to create, modify, and delete mapping rules while maintaining coordination among gNB groups. The system facilitates AI/ML-driven energy saving actions by enabling gNBs to interpret and compare energy cost information from neighboring nodes without additional conversion, supporting intelligent traffic offloading decisions to optimize overall network energy efficiency.