Sequential Feature Scheduling for Power-Efficient Sensor Classification
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Electronic devices face significant power consumption issues due to continuous processing of sensor data from multiple active sensors, leading to reduced battery life.
Innovation Solution
A device schedules features for sequential computing based on estimated power usage, evaluating a termination condition after each feature is computed to minimize unnecessary processing and conserve power.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If continuous processing of sensor data is performed, then accurate classification is maintained, but power consumption increases
Solution Approach 1:
The patent segments the feature processing into multiple stages, where simple features are processed first and complex features are processed only if needed. This segmentation allows the system to maintain classification accuracy by processing only necessary features, thereby reducing overall power consumption while preserving reliability.
Solution Approach 2:
The system performs partial processing by evaluating termination conditions after each feature computation. If the termination condition is met, further feature processing is stopped, meaning only the necessary portion of processing is performed. This partial action maintains sufficient classification accuracy while avoiding excessive power consumption from unnecessary computations.
2Reliability
If all features are computed, then classification accuracy is improved, but processing time increases
Solution Approach 1:
The patent applies preliminary action by computing simpler features first that can provide early classification results. The termination condition is evaluated after each feature, allowing the system to stop processing when sufficient information is obtained. This preliminary processing approach reduces processing time while maintaining classification accuracy by avoiding computation of all features.
Solution Approach 2:
The system performs only the necessary amount of processing by evaluating termination conditions after each feature computation. When the termination condition is met, processing stops, meaning only partial processing is performed rather than computing all features. This reduces processing time while maintaining sufficient classification accuracy.
3Productivity
If intensive processing is performed, then processing speed is improved, but power consumption increases
Solution Approach 1:
The patent implements dynamic processing by adjusting the level of feature computation based on termination conditions. The system starts with simpler, faster computations and only progresses to more intensive processing if needed. This dynamic approach maintains processing speed for simple cases while reducing power consumption by avoiding intensive processing when not necessary.
Solution Approach 2:
The system performs partial processing by evaluating termination conditions after each feature computation. When the condition is met, processing stops, meaning only the necessary portion is performed. This partial action maintains adequate processing speed for classification tasks while significantly reducing power consumption by avoiding unnecessary intensive computations.
Data Source
AI summary
Disclosed is an apparatus and method for power efficient processor scheduling of features. In one embodiment, features may be scheduled for sequential computing, and each scheduled feature may receive a sensor data sample as input. In one embodiment, scheduling may be based at least in part on each respective feature's estimated power usage. In one embodiment, a first feature in the sequential schedule of features may be computed and before computing a second feature in the sequential schedule of features, a termination condition may be evaluated.