AI Beam Prediction Pairing for Shared Wireless Measurements
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing AI-based beam management systems face unnecessary measurement overhead due to overlapping inputs and outputs of spatial and time domain beam prediction models, leading to inefficient resource utilization.
Innovation Solution
Implement a framework for matching and reusing beam measurement results between spatial and time domain AI models, reducing redundant measurements by ensuring input and output parameter compatibility.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If separate beam measurement processes are performed for spatial domain AI model and time domain AI model, then beam prediction accuracy is maintained, but measurement overhead increases
Solution Approach 1:
The patent merges the beam measurement processes of spatial domain AI model and time domain AI model into a unified measurement framework. By identifying that both models require beam measurement inputs, the invention combines these measurements so that a single measurement process serves both models simultaneously, eliminating redundant measurements while preserving the predictive accuracy of both spatial and temporal beam predictions
Solution Approach 2:
The patent creates a universal beam measurement mechanism that serves multiple functions - it provides measurement data for both spatial domain prediction and time domain prediction. The measurement process is designed to be multi-functional, extracting relevant information for different AI models from a single set of measurements, thereby reducing overall measurement overhead while maintaining comprehensive beam prediction capabilities
2Quantity of substance
If beam measurement results are reused between spatial and time domain AI models, then measurement overhead is reduced, but system complexity increases
Solution Approach 1:
The patent segments the beam management system into distinct functional modules: a measurement module that performs beam measurements, a result processing module that manages measurement results, and multiple AI model modules (spatial and time domain) that consume these results. This segmentation allows for organized result sharing through defined interfaces and data structures, reducing system complexity despite the increased functionality of result reuse
Solution Approach 2:
The patent introduces an intermediary result management mechanism that mediates between the measurement process and multiple AI models. This intermediary layer standardizes how measurement results are stored, accessed, and shared between different models, providing a structured approach that simplifies the complexity of result reuse by establishing clear data flow patterns and access protocols
Data Source
Figure 1~2
Figure 3~4
Figure 5~6
AI summary
The present disclosure provides an electronic device and method on a user equipment side and a base station side for wireless communication, and a computer readable storage medium. The electronic device on the user equipment side for wireless communication comprises: a processing circuit configured to: pair a first artificial intelligence model used for performing spatial domain beam prediction and a second artificial intelligence model used for performing time domain beam prediction; and multiplex a beam measurement result between the first artificial intelligence model and the second artificial intelligence model if the pairing is successful.