Channel Estimation Model Selection for Varying Signal Quality

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Existing channel estimation methods, particularly in hybrid beamforming systems, face challenges in achieving accurate channel estimation due to varying signal qualities and computational complexity, which affects data recovery and beam alignment.

Innovation Solution

A machine learning-based approach that dynamically selects a channel estimation model based on the expected signal quality, training multiple models for different candidate signal qualities to improve accuracy and reduce computational complexity.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Measurement precision

If multiple channel estimation models are trained for different signal qualities, then channel estimation accuracy is improved, but device complexity increases

Engineering Contradiction:
Improvechannel estimation accuracyVSAvoidmodel selection complexity
Core Design Contradiction:
Measurement precisionVSDevice complexity

Solution Approach 1:

The system dynamically selects different channel estimation models based on real-time signal quality conditions. Instead of using a single fixed model, the apparatus adapts the model selection to match current channel conditions, thereby maintaining high estimation accuracy across varying signal qualities without permanently increasing system complexity

Inventive Principle:
Principle #15Dynamics

Solution Approach 2:

The system changes the parameter of model selection based on signal quality measurements. By monitoring signal quality metrics and selecting appropriate pre-trained models corresponding to different signal quality ranges, the system achieves accurate channel estimation for both good and poor channel conditions while managing complexity through parameter-based model selection

Inventive Principle:
Principle #35Parameter changes

2Device complexity

If a single channel estimation model is used, then device complexity is reduced, but channel estimation accuracy deteriorates under varying signal qualities

Engineering Contradiction:
Improvemodel selection complexityVSAvoidchannel estimation accuracy
Core Design Contradiction:
Device complexityVSMeasurement precision

Solution Approach 1:

The system segments the channel estimation task into multiple specialized models, each trained for specific signal quality ranges. This segmentation allows each model to be optimized for particular conditions, improving overall accuracy while the segmentation structure itself provides a manageable complexity framework through clear division of labor among models

Inventive Principle:
Principle #1Segmentation

3Measurement precision

If channel estimation is performed with high computational complexity, then channel estimation accuracy is improved, but processing latency increases

Engineering Contradiction:
Improvechannel estimation accuracyVSAvoidprocessing latency
Core Design Contradiction:
Measurement precisionVSLoss of time

Solution Approach 1:

Multiple channel estimation models are pre-trained for different signal quality conditions before actual channel estimation is needed. This preliminary action allows the system to have ready-to-use models that can be quickly selected and applied during real-time operation, avoiding the time-consuming process of training models on-the-fly while maintaining high estimation accuracy

Inventive Principle:
Principle #10Preliminary action

Solution Approach 2:

The system dynamically selects from pre-trained models based on current signal quality, enabling fast adaptation without retraining. This dynamic selection approach maintains high accuracy by matching models to conditions while reducing latency compared to single-model approaches that require more complex processing to handle varying conditions

Inventive Principle:
Principle #15Dynamics

Data Source

PatentEP4032341B1Machine learning-based channel estimation
Publication Date: 2026.01.14 NOKIA SOLUTIONS & NETWORKS OY
  • EP4032341B1 patent drawingFigure 1A~1B
  • EP4032341B1 patent drawingFigure 2~3
  • EP4032341B1 patent drawingFigure 4

AI summary

Example embodiments of the present disclosure relate to machine learning-based channel estimation. According to an example embodiment, a first device determines a signal quality that is expected in transmission of a reference signal from a second device to the first device and receives the reference signal from the second device. The first device selects, based on the expected signal quality, a channel estimation model from a plurality of channel estimation models that have been trained for a plurality of candidate signal qualities for the reference signal. The first device determines, using the selected channel estimation model and based on the received reference signal, channel state information of a communication channel from the first device to the second device. According to this solution, a channel estimation model is dynamically selected for use, depending on a real-time signal quality expected to be gained in transmission of a certain RS.