Wireless AI Model Switching via Configuration Message Activation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing wireless communication systems lack efficient methods for managing and activating/deactivating AI/ML models, leading to suboptimal performance and resource inefficiencies.
Innovation Solution
Implementing a system for AI/ML model management and activation/deactivation through configuration messages that include activation fields, allowing wireless devices and network devices to exchange capability information and activate specific AI models based on conditions and requirements.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If AI/ML models are continuously running in wireless communication systems, then performance and accuracy are improved, but power consumption and resource usage increase
Solution Approach 1:
The patent implements dynamic activation and deactivation of AI/ML models based on real-time conditions. The system monitors traffic patterns, channel conditions, and device states to selectively activate models only when needed, rather than running them continuously. This dynamic approach maintains prediction accuracy when conditions warrant model usage while significantly reducing power consumption during low-activity periods.
Solution Approach 2:
The system employs periodic evaluation of performance metrics and resource usage to determine when AI models should be activated or deactivated. By periodically assessing whether the benefits of AI processing outweigh the energy costs, the system can switch between AI-based and traditional processing modes, optimizing the balance between accuracy and power consumption over time.
2Adaptability or versatility
If multiple AI models are deployed to handle different scenarios, then adaptability and versatility are improved, but device complexity increases
Solution Approach 1:
The patent implements a unified model management framework that handles multiple AI models through a single standardized interface and control mechanism. This universal management system can activate, deactivate, and switch between different AI models without requiring separate management logic for each model, thereby supporting scenario coverage while controlling overall system complexity.
Solution Approach 2:
The system segments the AI model management functionality into distinct modular components: model selection logic, activation control, performance monitoring, and switching mechanisms. This segmentation allows each component to be independently optimized and managed, reducing the complexity burden of handling multiple AI models while maintaining comprehensive scenario coverage.
3Speed
If AI models are activated based on frequent condition changes, then responsiveness and performance are improved, but signaling overhead and processing time increase
Solution Approach 1:
The system implements threshold-based activation criteria that require conditions to exceed certain levels before triggering AI model activation or deactivation. This partial action approach filters out minor fluctuations in channel conditions or traffic patterns, reducing the frequency of model switching while still responding promptly to significant changes that truly warrant model transitions.
Data Source
AI summary
Apparatuses, systems, and methods for AI/ML model management and activation/deactivation. A wireless device comprising at least one antenna and a processor is configured to: receive a configuration message from a network device, the configuration message comprising indication information of a list of Artificial Intelligence (AI) models, wherein the configuration message further comprises an activation field indicating a first AI model in the list of AI models to be activated; and activate the first AI model indicated by the activation field.