Wireless Device ML Fallback Features for Robust Operation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current wireless communication systems lack mechanisms to ensure robustness and resilience of machine learning models deployed at user equipment (UE) for critical functionalities, leading to potential performance issues such as incorrect model outputs affecting network decisions and impacting communication performance.
Innovation Solution
User equipment (UE) is equipped with both machine learning (ML)-based features and fallback features for critical functionalities, enabling it to switch to fallback features when performance issues are detected, and communicate this capability to the network.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If machine learning models are deployed at user equipment for critical functionalities, then performance and accuracy are improved, but reliability deteriorates when model outputs become incorrect
Solution Approach 1:
The patent implements a fallback feature that is prepared in advance to compensate for potential ML model failures. When the ML model produces incorrect outputs or performance degradation is detected, the system automatically switches to the pre-configured fallback feature, ensuring continuous reliable operation of critical functionalities without service interruption
Solution Approach 2:
The system dynamically changes operational parameters by switching between different feature sets (ML-based features vs. fallback features) based on model performance. This parameter change allows the system to adapt to varying conditions, maintaining reliability by selecting the appropriate feature set depending on whether the ML model is functioning correctly or has degraded
2Reliability
If fallback features are added to ensure robustness, then reliability is improved, but device complexity increases
Solution Approach 1:
The fallback feature is designed to provide the same critical functionality as the ML-based feature, making the system multi-functional. This universal approach allows a single system to handle both ML-based operation and fallback operation, reducing the need for entirely separate systems and thereby limiting the increase in device complexity while maintaining reliability
3Reliability
If the system switches to fallback features when ML performance degrades, then reliability is maintained, but loss of time occurs during switching
Solution Approach 1:
The fallback feature is pre-configured and prepared in advance, so when switching is needed, the system can activate it immediately without extensive setup or computation time. This preliminary preparation minimizes switching latency and ensures that reliability is maintained with minimal time loss during the transition from ML-based to fallback operation
Data Source
AI summary
According to some embodiments, a method is performed by a wireless device for fallback operation of a machine learning (ML) model. The method comprises: transmitting a message indicating a capability of the wireless device for supporting a combination of at least one ML-based feature for a functionality and at least one fallback feature for the functionality to a network node; operating the at least one ML-based feature for the functionality; and operating the at least one fallback feature for the functionality.