Dynamic Governing Metric Generator for Adaptive Resource Provisioning
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional metrics for optimizing electronic systems are inflexible and fail to adapt to changing high-level goals, leading to suboptimal performance in maintaining energy efficiency and other objectives.
Innovation Solution
A system and method for dynamically generating governing metrics that adjust to varying high-level goals over time and across different zones within a provisioned system, using a governing metric generator that includes input, low-level policy, property identification, metric formulation, monitoring, and output modules to optimize resource provisioning.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If conventional metrics are defined a priori for specific purposes, then they can achieve high-level goals initially, but they fail to adapt when goals change
Solution Approach 1:
The patent implements dynamic metric generation where metrics are continuously created and updated based on current system state and changing goals, rather than using static pre-defined metrics. The metric generation module dynamically formulates metrics that adapt to varying high-level goals, ensuring both reliability in goal achievement and adaptability to changing conditions.
Solution Approach 2:
The system changes the parameters of metrics based on the current operating context and goal priorities. By dynamically adjusting metric parameters according to changing system conditions and goal importance, the system maintains reliable goal achievement while adapting to new requirements without being constrained by fixed pre-defined metrics.
2Device complexity
If metrics are defined exclusively for a specific purpose, then they are simple and focused, but they become useless when the high-level goal changes
Solution Approach 1:
The patent creates a universal metric generation system that can produce metrics for multiple different high-level goals using a common framework. The metric generation module is designed to handle various goal types (energy efficiency, performance, cost, etc.) through a unified dynamic metric formulation approach, making the system versatile without requiring separate complex metric definitions for each goal.
Solution Approach 2:
By making metrics dynamic rather than static, the system achieves versatility across different goals without increasing inherent complexity. The same dynamic metric generation mechanism adapts to different goals by changing metric formulations based on current requirements, maintaining simplicity in the underlying system while achieving broad applicability.
3Productivity
If conventional metrics are used, then resource provisioning can be controlled initially, but optimization effectiveness decreases as conditions change
Solution Approach 1:
The patent implements continuous feedback loops where system performance is monitored against dynamically generated metrics, and metric formulations are updated based on feedback from changing conditions and goal achievement levels. This feedback mechanism ensures that resource provisioning remains effective and optimized even as system conditions evolve, preventing the degradation of optimization effectiveness over time.
Solution Approach 2:
The system uses dynamic metric adjustment to maintain high resource provisioning effectiveness as conditions change. By continuously updating metrics based on current system state and goal priorities, the system ensures that optimization remains effective throughout operation rather than degrading as conventional static metrics would.
Data Source
AI summary
In a method of generating governing metrics, a high-level goal to be met in a provisioned system is identified. In addition, a low-level governing policy designed to facilitate achievement of the high-level goal is selected and properties relating to the selected low-level governing policy are identified. The identified properties are formulated to define governing metrics relevant to the selected low-level governing policy and the formulated governing metrics are outputted. The formulated governing metrics are configured to be used in at least one of evaluating and controlling resource provisioning in the provisioned system.