ML Control Channel Resource Selection to Cut Blind Decoding

Solution Overview

Problem

Existing wireless communications systems face challenges in reducing power consumption and signaling overhead due to blind decoding processes in control channel resource selection, which increase end-to-end receive timelines and decoder dimensioning requirements.

Innovation Solution

Implementing machine learning-based techniques for control channel resource selection, utilizing neural networks trained by base stations to predict resource locations and aggregation levels, reducing the number of blind decoder attempts and optimizing decoding processes.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Reliability

If blind decoding processes are used for control channel resource selection, then all possible resource locations are monitored ensuring reliable reception, but power consumption increases and end-to-end receive timelines are extended

Engineering Contradiction:

Improvecontrol channel reception reliabilityVSAvoidpower consumption

Core Design Contradiction:

ReliabilityVSUse of energy by moving object

Solution Approach 1:

The network entity performs preliminary actions by training the ML model and generating predictions for control channel resource locations before the UE needs to decode. The UE receives these predictions in advance and uses them to directly identify resource locations without performing blind decoding, thereby reducing power consumption while maintaining reliable reception through accurate ML-based predictions

Inventive Principle:

Principle #10Preliminary action

Solution Approach 2:

The patent replaces the mechanical blind decoding process with an ML-based prediction system. Instead of systematically trying all possible resource locations (mechanical approach), the system uses a trained neural network to predict the most likely resource locations (intelligent substitution), significantly reducing the computational burden and power consumption at the UE

Inventive Principle:

Principle #28Mechanics substitution (Replace mechanical system)

2Reliability

If blind decoding processes are used for control channel resource selection, then all possible resource locations are monitored ensuring complete coverage, but end-to-end receive timelines are extended

Engineering Contradiction:

Improvecontrol channel reception reliabilityVSAvoidend-to-end receive timeline

Core Design Contradiction:

ReliabilityVSLoss of time

Solution Approach 1:

The network entity performs preliminary actions by training the ML model and generating predictions for control channel resource locations before the UE needs to decode. The UE receives these predictions in advance and uses them to directly identify resource locations without performing blind decoding, thereby reducing power consumption while maintaining reliable reception through accurate ML-based predictions

Inventive Principle:

Principle #10Preliminary action

Solution Approach 2:

The ML model allows the UE to skip the time-consuming blind decoding process entirely. By rushing through to the predicted resource locations directly based on ML predictions, the system achieves faster reception timelines while maintaining reliability through the accuracy of the trained model

Inventive Principle:

Principle #21Skipping (Rushing through)

3Reliability

If traditional control channel resource selection is used, then all possible aggregation levels are monitored ensuring complete coverage, but decoder dimensioning requirements increase

Engineering Contradiction:

Improvecontrol channel reception reliabilityVSAvoiddecoder dimensioning

Core Design Contradiction:

ReliabilityVSDevice complexity

Solution Approach 1:

The network entity performs preliminary actions by training the ML model and generating predictions for control channel resource locations before the UE needs to decode. The UE receives these predictions in advance and uses them to directly identify resource locations without performing blind decoding, thereby reducing power consumption while maintaining reliable reception through accurate ML-based predictions

Inventive Principle:

Principle #10Preliminary action

Solution Approach 2:

The patent replaces the mechanical blind decoding process with an ML-based prediction system. Instead of systematically trying all possible resource locations (mechanical approach), the system uses a trained neural network to predict the most likely resource locations (intelligent substitution), significantly reducing the computational burden and power consumption at the UE

Inventive Principle:

Principle #28Mechanics substitution (Replace mechanical system)

Data Source

PatentUS12580794B2Machine learning based control channel resource selection

Publication Date: 2026.03.17 QUALCOMM INC

AI summary

Certain aspects of the present disclosure provide techniques for machine learning (ML) based control channel (CCH) resource selection. An example method, performed at a user equipment (UE), generally includes receiving, from a network entity, signaling indicating parameters for configuring a machine learning (ML) model, applying the parameters to configure the ML model, performing channel estimation based on at least one reference signal (RS) measurement, using the ML model to select control channel (CCH) resources to monitor, based on the channel estimation, and monitoring the selected CCH resources for a CCH transmission from the network entity.