Conditional AI Model Selection for Missing Wireless Input Data
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
In wireless communications systems, user equipment (UE) may use machine learning models that are prone to errors due to missing or corrupt input data, leading to incorrect predictions and high error rates.
Innovation Solution
Implement conditional machine learning model and parameter set configurations, where network entities configure model identifiers and parameter sets with usage conditions, allowing UEs to select appropriate models and sets based on available data, thereby managing missing or corrupt input data.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Extent of automation
If machine learning models are used for predictions in wireless communications, then prediction capability is improved, but error rate increases due to missing or corrupt input data
Solution Approach 1:
The patent changes the parameter state by introducing conditional configurations that adapt the machine learning model's input parameters based on data availability. When input data is missing or corrupt, the system switches to alternative parameter sets or fallback models, thereby maintaining reliability while preserving automation capability
Solution Approach 2:
The patent introduces an intermediary mechanism (the network entity) that mediates between the data source and the machine learning model. This intermediary validates input data, detects missing or corrupt parameters, and selectively provides cleaned or alternative input data to the model, reducing error rates while maintaining automated predictions
2Reliability
If multiple machine learning models are configured with different usage conditions, then reliability is improved by managing missing data, but device complexity increases
Solution Approach 1:
The patent segments the machine learning solution into multiple conditional models or parameter sets, each optimized for specific data availability scenarios. This segmentation allows the system to select the appropriate model based on current conditions, improving reliability while managing complexity through modular organization
Solution Approach 2:
The patent implements dynamic model selection where the system automatically adjusts which machine learning model or parameter set is active based on real-time data availability conditions. This dynamic approach maintains high reliability across varying conditions while keeping the overall system manageable through automated adaptation
Data Source
AI summary
Methods, systems, and devices for wireless communications are described. A user equipment (UE) may receive a message from a network entity indicating a set of machine learning models, a set of parameter sets, or both and one or more usage conditions associated with the machine learning models and parameter sets. Based on a usage condition being satisfied, the UE may select a machine learning model, a parameter set, or both for generating a machine learning inference. For example, the UE may select the machine learning model or the parameter set based on a priority, whether sufficient input data is provided, or based on other usage conditions. The UE may generate the machine learning inference using the selected machine learning model or the selected parameter set, and the UE may transmit a report indicating an output of the machine learning inference to the network entity.