Cellular AI Training Data Alignment Using UE and BS Metrics
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing wireless communication systems face challenges in aligning data from user equipment (UE) and base stations (BS) for effective AI/ML model training due to misaligned clocks and missing metrics, which hinders efficient data preparation for 5G/NR systems.
Innovation Solution
A method and apparatus for segmenting UE and BS data based on common metrics and aligning them in the time domain to ensure synchronized data alignment for AI/ML model training, using techniques such as down-sampling and selecting common alignment metrics like MCS and ACK/NACK for precise alignment.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Quantity of substance
If data from UE and BS are collected for AI/ML model training, then training data availability is improved, but data alignment accuracy deteriorates due to misaligned clocks and missing metrics
Solution Approach 1:
The patent segments the collected data from UE and BS into multiple time-aligned segments based on common metrics (MCS, ACK/NACK). This segmentation allows the data to be divided into manageable chunks that can be precisely aligned in time domain, resolving the contradiction by enabling both sufficient data quantity for training and accurate time synchronization through segment-based alignment.
Solution Approach 2:
The patent introduces common metrics (MCS, ACK/NACK) as intermediary elements to bridge the UE and BS data streams. These common metrics serve as synchronization markers that enable accurate time alignment between the two data sources, allowing the system to maintain data alignment accuracy while collecting comprehensive data for training.
2Measurement precision
If data segmentation and alignment processes are implemented, then data alignment accuracy is improved, but processing complexity increases
Solution Approach 1:
The patent implements self-service alignment where the data segmentation and alignment process automatically utilizes the common metrics (MCS, ACK/NACK) that are already present in the data streams. The system leverages existing data structures and metrics to perform alignment without requiring external synchronization signals or complex manual configuration, thereby reducing processing complexity while maintaining accuracy.
Solution Approach 2:
The patent changes the alignment parameter from continuous time synchronization to discrete metric-based alignment. By using common metrics (MCS, ACK/NACK) as alignment anchors rather than continuous timestamps, the system simplifies the alignment process and reduces computational complexity while achieving accurate data synchronization for training.
Data Source
AI summary
Apparatuses and methods of data preparation for artificial intelligence (AI)/machine learning (ML) model training in cellular systems. A method includes receiving, by a network device, user equipment (UE) data from a UE and base station (BS) data from a BS; segmenting, based on a condition, the UE data and the BS data into a plurality of UE data segments and a plurality of BS data segments, respectively; identifying a common metric used in both a UE data segment of the plurality of UE data segments and a BS data segment of the plurality of BS data segments; and aligning the UE data segment with the BS data segments in time domain based on the common metric.