Interference-Aware AI/ML Beam Prediction in Wireless Communication

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Current AI/ML model-based beam (pair) prediction in wireless communication systems fails to reflect interference between beams, leading to suboptimal performance in beam management systems.

Innovation Solution

A wireless communication method that inputs measured power information of reference signals and interference noise into an AI/ML model to predict spatial filter identifiers and/or Layer1 Signal to Interference plus Noise Ratios (L1-SINRs), thereby reflecting interference between beams and improving beam management performance.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Reliability

If AI/ML model-based beam (pair) prediction is used, then beam management performance is improved, but interference between beams cannot be reflected

Engineering Contradiction:
Improvebeam management performanceVSAvoidinterference information
Core Design Contradiction:
ReliabilityVSLoss of information

Solution Approach 1:

The received signal power is segmented into two distinct components: reference signal power and interference noise power. This segmentation allows the AI/ML model to process interference information separately and accurately, resolving the contradiction by enabling interference reflection while maintaining beam management performance improvement.

Inventive Principle:
Principle #1Segmentation

Solution Approach 2:

The patent performs preliminary measurement and separation of reference signal power and interference noise power before feeding data into the AI/ML model. This preliminary action ensures that interference information is captured and processed in advance, allowing the model to reflect interference while still achieving performance improvement through predictive capabilities.

Inventive Principle:
Principle #10Preliminary action

2Area of stationary object

If multiple beams (beam pairs) are used to cover different terminals, then coverage is improved, but interference between beams increases

Engineering Contradiction:
Improvecoverage areaVSAvoidinterference between beams
Core Design Contradiction:
Area of stationary objectVSObject-affected harmful factors

Solution Approach 1:

The patent introduces feedback mechanisms where the AI/ML model predicts L1-SINR values that reflect interference levels. This feedback information is used to adjust beam selection and management decisions, allowing the system to maintain wide coverage while compensating for interference effects through informed decision-making.

Inventive Principle:
Principle #23Feedback

Solution Approach 2:

The patent changes the parameter set fed into the AI/ML model to include separately measured reference signal power and interference noise power. This parameter change enables the model to account for interference when making beam management decisions, allowing multiple beams to be used for coverage while managing interference through predictive analysis.

Inventive Principle:
Principle #35Parameter changes

Data Source

PatentEP4642088A1Wireless communication method and device
Publication Date: 2025.10.29 GUANGDONG OPPO MOBILE TELECOMMUNICATIONS CORP LTD
  • EP4642088A1 patent drawingFigure 1~3
  • EP4642088A1 patent drawingFigure 4~5
  • EP4642088A1 patent drawingFigure 6~7

AI summary

Provided in the embodiments of the present application are a wireless communication method and a device. A first communication device may input into a first network model power information of a reference signal portion and power information of an interference noise portion that are obtained by means of measurement, so as to perform prediction to obtain identification information of K spatial filters and/or L1-SINRs corresponding to the K spatial filters. That is, interference between beams (beam pairs) can be reflected in beam (pair) prediction based on an AI/ML model, thereby improving the performance of a beam management system.