Edge Resource Demand Load Scheduling via Predictive Estimation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Edge systems face challenges in managing machine-to-machine workloads due to limited resources and random workload demands, leading to inefficiencies in resource allocation and scheduling, particularly in geographically distributed environments with high uncertainty and limited access to real-time resource usage data.
Innovation Solution
The implementation of a predictive resource demand load estimation framework using Multivariate Probabilistic Collocation Method-Orthogonal Fractional Factorial Design (M-PCM-OFFD) to accurately estimate resource consumption and utilization, enabling efficient scheduling and resource allocation across edge systems.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If traditional scheduling methods are used in edge systems, then device complexity is reduced, but resource allocation efficiency deteriorates due to limited resources and random workload demands
Solution Approach 1:
The patent implements predictive resource demand load estimation that forecasts future resource requirements before workloads arrive at edge systems. By using machine learning models to predict demand patterns in advance, the system can proactively allocate resources and adjust scheduling decisions, transforming reactive scheduling into proactive resource management that improves allocation efficiency without requiring complex real-time coordination
Solution Approach 2:
The patent establishes a feedback mechanism where resource usage data from edge systems is continuously collected, analyzed, and fed back into the scheduling framework. This feedback loop enables the system to learn from actual resource consumption patterns, refine demand predictions, and dynamically adjust scheduling policies, creating an adaptive system that improves efficiency over time while maintaining manageable complexity through data-driven decision making
2Measurement precision
If real-time resource monitoring is implemented across distributed edge systems, then resource utilization accuracy is improved, but loss of time increases due to data collection and transmission delays
Solution Approach 1:
The patent implements predictive modeling that estimates future resource demand based on historical patterns and current trends, eliminating the need to wait for real-time data collection from all distributed edge systems. By forecasting resource requirements in advance, the system achieves accurate resource utilization assessment without the time delays associated with gathering and transmitting data from geographically distributed sources
Solution Approach 2:
The patent introduces a centralized orchestration layer that acts as an intermediary between edge systems and the scheduling framework. This intermediary aggregates resource data from multiple edge systems, applies predictive analytics, and generates scheduling decisions, reducing the time required for data collection and transmission while maintaining accurate resource utilization measurements through intelligent data synthesis
3Productivity
If predictive resource demand estimation is implemented, then resource allocation efficiency is improved, but device complexity increases due to the estimation framework
Solution Approach 1:
The patent implements a universal predictive estimation framework that uses machine learning models to handle multiple types of workloads and resource types across diverse edge systems. By creating a multi-functional prediction engine that can estimate demand for compute, storage, network, and various workload types using unified algorithms, the system improves resource allocation efficiency across the entire edge infrastructure without requiring separate complex estimation systems for each resource type
Solution Approach 2:
The patent employs parameter-based predictive modeling where resource demand is estimated by analyzing changes in key parameters such as workload arrival rates, resource consumption patterns, and system utilization metrics. By transforming complex multidimensional resource demand predictions into manageable parameter changes and trends, the system achieves accurate forecasting while keeping the estimation framework complexity controlled through dimensionality reduction and pattern recognition
4Adaptability or versatility
If edge systems handle machine-to-machine workloads with random demands, then adaptability is improved, but productivity deteriorates due to uncertainty in resource utilization
Solution Approach 1:
The patent implements dynamic scheduling policies that automatically adjust resource allocation based on real-time workload characteristics and predicted demand patterns. By making the scheduling system dynamic and adaptive to changing workload conditions rather than using static allocation rules, the system can handle random machine-to-machine workloads efficiently, improving both adaptability to different workload types and productivity through optimized resource utilization under uncertainty
Data Source
AI summary
Managing the resource demand load for edge systems is significantly more complex than for other systems, such as cloud environments. Edge resource demand load scheduling systems and methods are disclosed that can ensure that edge systems operate smoothly and efficiently while balancing multiple scheduling objectives. Scheduling techniques disclosed herein may utilize heuristic rules for candidate edge system selection (e.g., utilizing ARMA/ARIMA averages and/or service level objectives) and modified best fit decreasing (mBFD) assignment/allocation techniques.