Workload Shifting Policy Verification for Fair Edge Computing
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
The challenge of ensuring consistent and desired workload shifting in distributed cloud environments is difficult due to the complexity of AI-based learning models and resource-intensive operations, leading to excessive power consumption and processing bottlenecks.
Innovation Solution
A computer-implemented method and system for determining the fairness of workload shifting policies by replicating the policy through execution logs, using a probabilistic model checker to evaluate desired metrics and produce a quantitative measure of fairness.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If AI-based learning models are used to manage workloads, then workload shifting capability is improved, but model complexity and processing requirements increase
Solution Approach 1:
The patent creates a simplified copy of the complex AI-based workload shifting policy as a formal model. This model replicates the policy's behavior using standardized constructs (states, transitions, probabilities) that can be systematically analyzed without requiring the full complexity of the original AI model, thus enabling verification while reducing processing requirements.
Solution Approach 2:
The patent replaces the complex AI-based decision-making mechanism with a formal probabilistic model that uses mathematical constructs (states, transitions, probability distributions) to represent and analyze workload shifting behavior. This substitution enables rigorous analysis of policy fairness and performance without the computational burden of the original AI models.
2Productivity
If workloads are distributed across multiple cloud locations, then processing capacity is improved, but ensuring consistent and desired workload distribution becomes difficult
Solution Approach 1:
The patent implements a verification system that provides feedback on workload distribution fairness by analyzing execution logs against the formal model. The model checker compares actual workload distribution outcomes against the desired policy objectives, identifying discrepancies and providing measurable insights into distribution consistency across cloud locations.
Solution Approach 2:
The patent replaces complex AI-based workload distribution mechanisms with a formal probabilistic model that uses mathematical constructs (states, transitions, probability distributions) to represent and analyze workload shifting behavior. This substitution enables rigorous analysis of policy fairness and performance without the computational burden of the original AI models.
3Power
If resource intensive operations are performed, then processing capability is improved, but power consumption increases
Solution Approach 1:
The patent creates a simplified copy of the complex AI-based workload shifting policy as a formal model. This model replicates the policy's behavior using standardized constructs (states, transitions, probabilities) that can be systematically analyzed without requiring the full complexity of the original AI model, thus enabling verification while reducing processing requirements.
Solution Approach 2:
The patent changes the parameters of analysis from raw AI model outputs to formal probabilistic model parameters (state transition probabilities, expected values, fairness metrics). This parameter transformation enables energy-efficient analysis by working with simplified mathematical representations rather than the full complexity of the original AI models and execution logs.
Data Source
AI summary
A computer-implemented method according to one approach, is for determining fairness of a workload shifting policy. The CIM includes receiving execution logs from a system implementing the workload shifting policy, and inspecting the execution logs. A model is developed that replicates how the workload shifting policy is applied by the system. Moreover, a desired metric of interest is defined. A probabilistic model checker is used to evaluate the model and the desired metric of interest, and a quantitative measure of the fairness of the workload shifting policy is produced.