Beamforming Profile Adjustment Using Learning-Based Performance Prediction

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Existing methods for adjusting beamforming profiles in massive MIMO communication networks are inefficient and sub-optimal, particularly in dynamic environments with varying terminal device locations, traffic distribution, and interference, leading to prolonged manual optimization and inaccurate performance predictions.

Innovation Solution

Employing a machine learning method with a pre-trained learning model to predict beamforming performance based on historical measurement information, using reinforcement learning and artificial neural networks to dynamically adjust beamforming profiles for optimal performance.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Productivity

If manual optimization methods are used to adjust beamforming profiles, then implementation simplicity is maintained, but optimization efficiency is low and convergence to optimal profiles is prolonged

Engineering Contradiction:
Improveoptimization efficiencyVSAvoidconvergence time
Core Design Contradiction:
ProductivityVSLoss of time

Solution Approach 1:

The system employs a machine learning model that automatically adjusts beamforming profiles based on historical measurement information and performance feedback, enabling self-service optimization without manual intervention. The model continuously learns from network performance data and autonomously determines optimal beamforming configurations, resolving the contradiction by achieving both high efficiency and automatic operation.

Inventive Principle:
Principle #25Self-service

Solution Approach 2:

The patent replaces manual mechanical optimization processes with an intelligent machine learning system. The learning model uses algorithms to process measurement information and predict optimal beamforming profiles, substituting the slow manual adjustment mechanism with a rapid computational system that converges quickly to optimal solutions.

Inventive Principle:
Principle #28Mechanics substitution (Replace mechanical system)

2Adaptability or versatility

If traditional beamforming adjustment methods are used, then system complexity is low, but adaptability to dynamic environments with varying terminal locations and traffic distribution is poor

Engineering Contradiction:
Improveadaptability to dynamic environmentsVSAvoidsystem complexity
Core Design Contradiction:
Adaptability or versatilityVSDevice complexity

Solution Approach 1:

The beamforming profile adjustment system transitions from static manual configuration to dynamic automatic adaptation. The machine learning model continuously monitors changes in terminal device locations, traffic distribution, and interference conditions, automatically adjusting beamforming profiles in real-time to maintain optimal performance in dynamic environments.

Inventive Principle:
Principle #15Dynamics

Solution Approach 2:

The system implements a feedback mechanism where measurement information about network performance, terminal locations, and traffic patterns is continuously collected and fed back to the learning model. This feedback loop enables the system to adapt to changing conditions by learning from actual performance data and adjusting beamforming profiles accordingly, achieving high adaptability through intelligent feedback-driven optimization.

Inventive Principle:
Principle #23Feedback

3Measurement precision

If conventional performance prediction methods are used, then measurement accuracy is acceptable, but prediction precision for beamforming performance is inaccurate

Engineering Contradiction:
Improveperformance prediction accuracyVSAvoidprediction accuracy
Core Design Contradiction:
Measurement precisionVSLoss of information

Solution Approach 1:

The learning model is pre-trained using historical measurement information and performance data before actual beamforming optimization. This preliminary training phase allows the model to learn patterns and relationships in the data, improving its prediction accuracy when making real-time beamforming adjustments. The pre-training ensures the system has accurate prediction capabilities before deployment.

Inventive Principle:
Principle #10Preliminary action

4Productivity

If dynamic adjustment of beamforming profiles is implemented, then coverage and network performance are improved, but interference management and MU-MIMO pairing ratio may be negatively impacted

Engineering Contradiction:
Improvenetwork performanceVSAvoidinterference
Core Design Contradiction:
ProductivityVSObject-affected harmful factors

Solution Approach 1:

The system uses feedback from measurement information about interference levels and MU-MIMO pairing ratios to adjust beamforming profiles dynamically. When interference increases or pairing ratios deteriorate, the learning model modifies beamforming configurations to mitigate these harmful effects while maintaining overall network performance, achieving a balance between coverage improvement and interference management.

Inventive Principle:
Principle #23Feedback

Data Source

PatentEP3939367B1Adjusting beamforming profiles
Publication Date: 2026.02.11 NOKIA TECHNOLOGIES OY
  • EP3939367B1 patent drawingFigure 1
  • EP3939367B1 patent drawingFigure 2A~2B
  • EP3939367B1 patent drawingFigure 3

AI summary

Embodiments of the present disclosure relate to an electric device, a method for adjusting beamforming profiles and computer readable medium. The electric device comprises at least one processor and at least one memory including computer program code. The at least one memory and the computer program code are configured to, with the at least one processor, cause the electric device to: obtain performance measurement information for a broadcast area where an initial beamforming profile is currently used, the initial beamforming profile being determined from a set of candidate beamforming profiles; determine, based on the performance measurement information, the initial beamforming profile, each candidate beamforming profile in the set and a learning model, a respective beamforming performance estimation in case where a respective candidate beamforming profile is used in the broadcast area, wherein the learning model specifies an association between historical measurement information for the broadcast area and beamforming performance; and select one candidate beamforming profile from the set for the broadcast area based on the beamforming performance estimation and the performance measurement information such that an optimal beamforming performance can be obtained by using the selected candidate beamforming profile.