UE DRX Lifecycle Prioritization for ML Data Collection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing technologies do not effectively address the impact of discontinuous reception (DRX) cycles on data collection and machine learning (ML) life cycles in user equipment (UE), leading to disruptions and inefficiencies in power consumption and predictive accuracy, particularly in AI/ML-enabled mobility enhancements.
Innovation Solution
Implementing prioritization policies and RRC configuration parameters to synchronize and prioritize either DRX or ML functionalities based on collision conditions, allowing the UE to autonomously determine actions to enhance power savings or ML operations, such as skipping, extending, or triggering early wake-up signals during DRX cycles.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Use of energy by moving object
If DRX cycles are implemented to save power, then power consumption is reduced, but data collection for ML life cycle management is disrupted
Solution Approach 1:
The UE performs preliminary actions by collecting data during DRX active periods and storing it for later use during ML life cycle management operations. The network entity is notified in advance of data collection needs, allowing it to prepare necessary information before DRX inactive periods begin, ensuring continuous ML functionality despite power-saving interruptions.
Solution Approach 2:
The network entity acts as an intermediary between the UE's DRX power-saving mechanism and ML life cycle management requirements. It coordinates wake-up signals, transfers necessary data, and manages the timing of ML operations to bridge the gap between DRX cycles and continuous ML data collection needs.
2Use of energy by moving object
If DRX cycles are implemented to conserve energy, then energy efficiency is improved, but predictive accuracy of ML models deteriorates
Solution Approach 1:
The UE collects and stores data during DRX active periods in advance of when it needs to perform ML predictive operations. This preliminary data gathering ensures that sufficient training data is available when the UE wakes up, maintaining predictive accuracy while minimizing the duration of active reception periods.
Solution Approach 2:
The system implements periodic data collection during DRX active periods, followed by periodic ML life cycle management operations. This periodic action pattern allows the UE to balance continuous learning needs with energy-efficient sleep periods, maintaining model accuracy through regular updates rather than continuous operation.
3Reliability
If ML functionalities are prioritized during DRX cycles, then predictive accuracy is maintained, but power consumption increases
Solution Approach 1:
The system dynamically adjusts the priority between DRX and ML functionalities based on real-time conditions. The UE and network entity negotiate and reconfigure DRX parameters dynamically, extending active periods when ML operations are critical and shortening them when power saving is the primary goal, creating a flexible balance rather than a fixed priority.
Solution Approach 2:
The system changes DRX parameters such as cycle length, active period duration, and wake-up timing based on the current operational state and ML life cycle requirements. By adjusting these parameters dynamically, the system can allocate more time to ML operations when needed while maintaining overall power efficiency through optimized DRX configuration.
4Productivity
If DRX and ML life cycles are synchronized, then operational efficiency is improved, but system complexity increases
Solution Approach 1:
The UE provides feedback to the network entity about its DRX status, data collection progress, and ML operational needs. This feedback loop enables the network to adjust DRX configuration and ML timing in real-time, achieving synchronization through continuous communication rather than complex predetermined coordination.
Solution Approach 2:
The system segments the DRX and ML life cycle management into separate but coordinated functions. The UE handles data collection during DRX active periods, while the network entity manages ML operation scheduling. This segmentation allows each component to operate independently with simpler logic, reducing overall system complexity while achieving operational efficiency through coordination.
Data Source
AI summary
Systems, methods, apparatuses, and computer program products for periodic data collection for LCM functions and LCM signaling to allow UE to perform model-based and/or functionality-based LCM. One method may include a UE transmitting, to a network entity, at least one discontinuous reception, DRX, and life cycle management, LCM, monitoring configuration signaling request; receiving, in response to the at least one DRX and LCM monitoring configuration signaling request, at least one DRX and LCM monitoring configuration signaling response from the network entity; prioritizing, based on the at least one DRX and LCM monitoring configuration signaling response, at least one of a DRX functionality or a machine learning, ML, functionality; and transmitting, to the network entity, a DRX and LCM monitoring status report.