Beam Information Exchange for AI-Based Beam Prediction in 5G
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
In communication systems using massive antenna arrays for 5G NR, the large number of transmitting and receiving beams increases system load and latency, especially when predicting optimal beam pairs.
Innovation Solution
Implementing a method where network devices and terminal equipment exchange beam information, including the number and pattern of beams, to enable the training or selection of an appropriate AI model for predicting optimal beams.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If a massive antenna array is used to form shaped beams with greater gain, then propagation losses are overcome and system coverage is ensured, but the number of transmitting and receiving beams increases, leading to increased system load and latency
Solution Approach 1:
The patent segments the beam management process into two parts: actual beam measurements and AI-predicted beam measurements. Instead of measuring all M*N beam pairs, the system divides them into a first set of beams for actual measurement and a second set for AI prediction, thereby reducing the measurement overhead while maintaining coverage reliability
Solution Approach 2:
The patent introduces an AI model as an intermediary between actual beam measurements and final beam pair selection. The AI model predicts optimal beam pairs from a subset of measured beams, acting as a mediator that reduces the need for exhaustive measurements of all beam combinations
2Loss of information
If all M*N beam pairs are measured, then complete beam information is obtained, but system load and latency are greatly increased
Solution Approach 1:
The patent applies partial action by measuring only a subset of beam pairs (first set and second set) rather than all M*N beam pairs. The AI model then extrapolates the remaining beam information from this partial measurement set, reducing measurement overhead while maintaining sufficient beam information for optimal beam selection
Solution Approach 2:
The patent performs preliminary measurements on a selected subset of beam pairs before using the AI model to predict the remaining beam pairs. This preliminary action on a reduced set enables the system to obtain sufficient beam information without the overhead of measuring all possible beam combinations
3Productivity
If an AI model is used to predict optimal beam from measurement results of a small number of beams, then system load and latency are reduced, but the network device or terminal equipment does not know all information such as the number of beams at the transmitter, the number of beams at the receiver, and the pattern of the beams
Solution Approach 1:
The patent performs preliminary exchange of beam configuration information (number of transmit beams, number of receive beams, and beam patterns) between the network device and terminal equipment before AI model prediction. This preliminary action ensures that the AI model has access to necessary configuration information while still allowing the system to benefit from reduced measurements
Solution Approach 2:
The patent implements a feedback mechanism where the network device and terminal equipment exchange beam configuration information to ensure the AI model has the necessary context for accurate prediction. This feedback loop maintains information completeness while enabling the use of AI-based prediction to reduce system load and latency
Data Source
AI summary
A beam information transmission apparatus, applicable to a network device, includes: a first receiver receives first request information. The first request information being used to indicate the network device to transmit at least the number and/or pattern parameter information of downlink transmitted beams and/or the number and/or pattern parameter information of uplink received beams; and a first transmitter transmits at least the number and/or pattern parameter information of the downlink transmitted beams, and/or the number and pattern parameter information of uplink received beams.