Terminal CSI Reporting with AI-Based Full-CSI Prediction
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current radio communication technologies lack specific details for AI-aided beam management, hindering high-accuracy channel estimation and efficient resource use, which are essential for enhancing communication throughput and quality.
Innovation Solution
Implementing a terminal and base station with AI models that utilize complementary and reduced CSI feedback methods, allowing for accurate prediction of full CSI based on past feedbacks, thereby reducing communication overhead.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Device complexity
If AI-aided beam management is implemented without specific details definition, then device complexity is reduced, but measurement precision and manufacturing precision of channel estimation deteriorate
Solution Approach 1:
The patent segments the channel estimation process into multiple CSI reports with different granularities (first CSI report for wideband, second CSI report for subband). This segmentation allows the system to maintain high measurement precision through detailed subband reporting while reducing overall device complexity by using AI to predict and infer information from the coarser wideband report, thus resolving the contradiction between system complexity and channel estimation accuracy.
2Measurement precision
If full CSI is reported at all timings, then measurement precision is improved, but loss of information increases due to communication overhead
Solution Approach 1:
The patent applies preliminary action by having the terminal predict the second CSI (subband-level) based on the first CSI (wideband-level) report using AI techniques. This prediction is performed in advance before the actual second CSI report is generated, allowing the base station to reconstruct full CSI information with high precision while significantly reducing communication overhead, thus resolving the contradiction between measurement precision and information loss.
3Productivity
If AI models are used for CSI prediction, then productivity is improved through reduced overhead, but device complexity increases
Solution Approach 1:
The patent implements self-service by having the terminal autonomously perform AI-based CSI prediction and generation without requiring complex base station processing. The terminal's AI model predicts the second CSI report based on the first CSI report, effectively making the terminal self-sufficient in generating accurate channel state information. This approach improves communication productivity through reduced overhead while keeping base station complexity low, as the heavy AI processing is distributed to the terminal.
Data Source
AI summary
A terminal according to one aspect of the present disclosure includes a receiving section that receives one channel state information (CSI) report configuration, and a control section that performs control of transmitting first CSI at a first timing and second CSI being different from the first CSI at a second timing being different from the first timing, based on the one CSI report configuration. According to one aspect of the present disclosure, preferable channel estimation/use of resources can be implemented.