Mobile Throttle Control Using UEI Prediction for Thermal Response
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing throttle control mechanisms in mobile devices react slowly to temperature changes, leading to drastic performance downgrades and poor user experience due to predefined thermal thresholds that do not consider operating environments or user scenarios.
Innovation Solution
A throttle control method that collects input data, generates user experience indices using machine learning models, and selects optimal throttle control parameters based on these indices to improve precision and timing, considering various factors such as downlink/uplink throughput, voice quality, and latency.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Temperature
If reactive thermal management techniques are used to manage temperature, then thermal specifications are protected, but processor performance suffers drastic downgrade impacting user experience
Solution Approach 1:
The system performs preliminary actions by proactively adjusting processor performance before thermal thresholds are violated. The machine learning model predicts future thermal states and preemptively throttles processor performance to prevent overheating, rather than reacting after temperature limits are reached. This approach maintains user experience by avoiding drastic performance downgrades while still protecting thermal specifications.
2Temperature
If aggressive processor performance reduction is applied to drop temperature, then thermal threshold is satisfied, but user experience deteriorates due to slow response of temperature sensors
Solution Approach 1:
The system implements feedback by continuously monitoring temperature sensor readings and using machine learning models to predict thermal trends. The feedback loop analyzes the slow response characteristics of thermistors and compensates for this delay by predicting future temperature states. This allows the system to make informed performance adjustment decisions that maintain user experience while ensuring thermal threshold compliance.
Solution Approach 2:
The system changes parameters by transitioning from reactive threshold-based control to predictive probabilistic control. Instead of waiting for temperature thresholds to be violated, the machine learning model continuously evaluates thermal states and adjusts processor performance parameters proactively. This parameter transformation enables smoother performance transitions that preserve user experience while maintaining thermal compliance.
3Device complexity
If predefined thermal thresholds are used for throttle control, then thermal management is simplified, but control precision deteriorates by not considering operating environments or user scenarios
Solution Approach 1:
The system applies dynamics by transitioning from static predefined thermal thresholds to dynamic adaptive thresholds. The machine learning model continuously adapts thermal management parameters based on real-time operating conditions, user scenarios, and environmental factors. This dynamic approach maintains relatively simple device architecture while achieving high control precision by adjusting throttle parameters according to current system state and predicted thermal trends.
Data Source
AI summary
A throttle control method for a mobile device include collecting input data, generating a first set of user experience indices according to the input data, and checking whether a user experience index of the first set of user experience indices satisfies a UEI threshold. The input data includes common information data, current configuration data and a plurality of throttle control parameters. Each user experience index of the first set of user experience indices is corresponding to at least one of throttle control parameter of the plurality of throttle control parameters.