Selective Beam Disabling in Wireless Prediction Reports

Solution Overview

Problem

Existing wireless communication systems face inefficiencies in beam prediction due to unavailable or poorly performing beams, leading to resource consumption and latency when retraining or reconfiguring AI/ML models.

Innovation Solution

A UE and network node system that enables disabling specific beams in beam prediction models by indicating a subset of beams to be included or excluded from measurement reports, reducing the need for retraining or reconfiguration.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Measurement precision

If the AI/ML model is retrained or reconfigured to exclude unavailable beams, then the beam prediction accuracy is improved, but processing resources are consumed and latency is introduced

Engineering Contradiction:

Improvebeam prediction accuracyVSAvoidlatency

Core Design Contradiction:

Measurement precisionVSLoss of time

Solution Approach 1:

The network node pre-configures the AI/ML model with a superset of beams that includes both currently available beams and potentially unavailable beams. This preliminary configuration avoids the need for retraining when beams become unavailable, as the model is already trained on a comprehensive beam set. The UE receives configuration indicating which beams from the superset are currently available for reporting.

Inventive Principle:

Principle #10Preliminary action

Solution Approach 2:

The system dynamically adjusts the set of beams included in measurement reports based on real-time availability without requiring model reconfiguration. The network node can update the beam configuration to reflect current availability status, allowing the system to adapt to changing conditions (beam availability, traffic load, energy saving mode) without the overhead of retraining the AI/ML model.

Inventive Principle:

Principle #15Dynamics

2Measurement precision

If the AI/ML model is retrained or reconfigured to exclude unavailable beams, then the beam prediction accuracy is improved, but processing resources are consumed

Engineering Contradiction:

Improvebeam prediction accuracyVSAvoidprocessing resources

Core Design Contradiction:

Measurement precisionVSUse of energy by moving object

Solution Approach 1:

The network node pre-configures the AI/ML model with a superset of beams that includes both currently available beams and potentially unavailable beams. This preliminary configuration avoids the need for retraining when beams become unavailable, as the model is already trained on a comprehensive beam set. The UE receives configuration indicating which beams from the superset are currently available for reporting.

Inventive Principle:

Principle #10Preliminary action

Solution Approach 2:

The system dynamically adjusts the set of beams included in measurement reports based on real-time availability without requiring model reconfiguration. The network node can update the beam configuration to reflect current availability status, allowing the system to adapt to changing conditions (beam availability, traffic load, energy saving mode) without the overhead of retraining the AI/ML model.

Inventive Principle:

Principle #15Dynamics

3Measurement precision

If the AI/ML model is retrained or reconfigured to exclude unavailable beams, then the beam prediction accuracy is improved, but network resources are consumed

Engineering Contradiction:

Improvebeam prediction accuracyVSAvoidnetwork resources

Core Design Contradiction:

Measurement precisionVSQuantity of substance

Solution Approach 1:

The network node pre-configures the AI/ML model with a superset of beams that includes both currently available beams and potentially unavailable beams. This preliminary configuration avoids the need for retraining when beams become unavailable, as the model is already trained on a comprehensive beam set. The UE receives configuration indicating which beams from the superset are currently available for reporting.

Inventive Principle:

Principle #10Preliminary action

Solution Approach 2:

The system dynamically adjusts the set of beams included in measurement reports based on real-time availability without requiring model reconfiguration. The network node can update the beam configuration to reflect current availability status, allowing the system to adapt to changing conditions (beam availability, traffic load, energy saving mode) without the overhead of retraining the AI/ML model.

Inventive Principle:

Principle #15Dynamics

Data Source

PatentUS12432575B2Disabling beam prediction outputs

Publication Date: 2025.09.30 QUALCOMM INC

AI summary

Various aspects of the present disclosure generally relate to wireless communication. In some aspects, a user equipment (UE) may receive, from a network node, a configuration of a beam prediction model that is trained to predict beam measurements for a set of beams. The UE may receive, from the network node, an indication of a subset of beams, from the set of beams, that are to be associated with a measurement report. The UE may transmit, to the network node, the measurement report indicating one or more predicted beam measurements that are based at least in part on an output of the beam prediction model, the output of the beam prediction model including beam predictions associated with the set of beams, and the one or more predicted beam measurements including information associated with the subset of beams. Numerous other aspects are provided.