5G Beam Measurement Reporting for Lower AI Prediction Latency
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
The large number of transceiving beams in 5G NR systems leads to increased payload and latency due to the need for measuring M*N beams, which is not effectively addressed by existing methods, especially when AI models are deployed for beam prediction.
Innovation Solution
An information transceiving apparatus is provided for both terminal equipment and network devices, enabling configuration of specific beam measurement reporting through reference signal sets, receiving beam sets, and transceiving beam pairs, allowing the terminal equipment to report measurement results as specified by the network device, thereby facilitating AI model operation in training, inference, and performance monitoring.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If the terminal equipment measures all M*N beams in legacy beam measurement, then the beam measurement coverage is complete, but the system payload and latency are greatly increased
Solution Approach 1:
The patent applies partial action by having the terminal equipment measure only a subset of beams (K beams) instead of all M*N beams. The network device configures the terminal to measure and report only K beam measurement results, where K is less than M*N. This partial measurement approach reduces the payload and latency while the network device uses AI models to predict the remaining beam measurements, achieving a balance between measurement completeness and system efficiency.
2Loss of information
If the terminal equipment measures and reports all beam measurement results, then the AI model can be trained with complete data, but the system payload is greatly increased
Solution Approach 1:
The patent extracts only the necessary beam measurement data (K beams) that needs to be reported by the terminal equipment, while the remaining beam measurement data is obtained through AI model prediction at the network device side. This extraction approach reduces the payload transmitted over the air interface while ensuring the AI model receives sufficient training data through the combination of actual measurements and predicted values.
3Extent of automation
If the network device configures the terminal to report measurement results for AI model operation, then the AI model can be deployed for beam prediction, but the terminal equipment needs clear guidance on which beams to report
Solution Approach 1:
The patent segments the beam measurement reporting process into configurable components. The network device sends reporting configuration information that includes first information about a first reference signal set, second information about a first receiving beam set, and third information about a set of first transceiving beam pairs. This segmentation allows flexible configuration of which specific beams the terminal should measure and report, making the AI model deployment process more manageable and less complex.
Data Source
AI summary
An information transceiving apparatus, applicable to a network device, includes: a transmitter configured to transmit reporting configuration information to a terminal equipment, the reporting configuration information comprising first information relating a first reference signal set, and/or second information relating a first receiving beam set, and/or third information relating a set of first transceiving beam pairs comprising first transmitting beams and first receiving beams; and a receiver configured to receive a measurement report transmitted by the terminal equipment, the measurement report comprising measurement results to which first reference signals in the first reference signal set correspond, and/or measurement results to which first receiving beams in the first receiving beam set correspond, and/or measurement results to which first transceiving beam pairs in the set of first transceiving beam pairs correspond.