Systems and methods for artificial intelligence / machine learning models for beam management spatial prediction
By using AI/ML models for spatial beam prediction in wireless communication systems, the overhead of radio resource management measurements is reduced, the problem of low efficiency in high-frequency communication is solved, and higher communication throughput and system performance are achieved.
CN122228629APending Publication Date: 2026-06-16APPLE INC
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- APPLE INC
- Filing Date
- 2024-11-20
- Publication Date
- 2026-06-16
AI Technical Summary
Technical Problem
In wireless communication systems, the overhead of radio resource management (RRM) measurements in existing technologies is relatively large, especially in high-frequency bands such as FR2, which leads to reduced communication efficiency.
Method used
Artificial intelligence (AI)/machine learning (ML) models are used for spatial beam prediction to reduce the Tx/Rx beam scan set. L3 measurement latency and overhead are reduced by periodically skipping actual measurements and using model predictions, thereby optimizing RRM performance.
Benefits of technology
By reducing the number of beam scans in the wireless communication system, communication throughput is improved, and RRM and overall system performance are enhanced.
✦ Generated by Eureka AI based on patent content.
Smart Images
Figure CN122228629A_ABST
Abstract
A user equipment (UE) uses an artificial intelligence (AI) model to generate predicted receive signal strengths for Tx-Rx beam pairs between base station transmit (Tx) beams and UE receive (Rx) beams; identifies, from the Tx-Rx beam pairs, a predicted best Tx-Rx beam pair having a highest predicted receive signal strength; indicates, to the base station, a predicted best Tx beam of the predicted best Tx-Rx beam pair; performs reference signal measurements using a corresponding paired predicted best Rx beam of the predicted best Tx-Rx beam pair by the predicted best Tx beam transmitted by the base station to generate a measured receive signal strength for the predicted best Tx-Rx beam pair; and calculates a model validity metric for the AI model based on a comparison of the highest predicted receive signal strength and the measured receive signal strength. Other implementations of the AI model alternatively located at the base station are also discussed.
Need to check novelty before this filing date? Find Prior Art