Edge Utility System Aggregating Distributed Computing Resources
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing information processing systems face challenges in efficiently managing resources to meet varying workload demands across multiple edge computing sites, particularly in ensuring adequate resource availability for a large number of users.
Innovation Solution
The dynamic aggregation of edge resources across multiple edge computing sites is facilitated through an edge utility system, where processing devices receive user inputs, populate data structures, and aggregate resources to form an edge network, enabling the execution of workloads based on user-defined preferences and priorities.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If edge resources are statically allocated at each edge computing site, then resource availability is predictable, but the system cannot meet varying workload demands across multiple sites
Solution Approach 1:
The patent merges edge resources from multiple geographically distributed edge computing sites into a unified virtualized resource pool. This consolidation enables dynamic allocation of compute, storage, and network resources across sites, allowing the system to adapt to varying workload demands while presenting a simplified management interface through centralized orchestration.
Solution Approach 2:
The system implements dynamic resource allocation where edge resources are not statically assigned but continuously adjusted based on real-time workload demands. The virtualization layer enables resources to be dynamically provisioned, allocated, and de-provisioned across multiple edge sites, transforming the static resource model into a flexible, demand-responsive system.
2Adaptability or versatility
If virtual resources are used to meet changing user needs, then flexibility is improved, but resource availability cannot be ensured for a large number of users
Solution Approach 1:
The patent creates a universal edge resource pool that serves multiple users and workloads simultaneously. The virtualized infrastructure enables a single pool of physical resources to be shared across numerous users, with each user receiving guaranteed resource availability through standardized allocation policies while maintaining flexibility to accommodate different workload types and user requirements.
3Quantity of substance
If edge resources are aggregated across multiple sites, then resource pool size increases, but system complexity increases
Solution Approach 1:
The patent introduces a virtualization layer and centralized orchestration system as intermediaries between the physical edge resources at multiple sites and the end users. This intermediary layer abstracts the complexity of resource aggregation, presenting a simplified interface for resource management while handling the complex tasks of resource pooling, allocation, and coordination across geographically distributed sites.
4Ease of operation
If distributed edge computing sites are used, then proximity to users is improved, but inadequate resources arise at individual sites
Solution Approach 1:
The patent extends the resource pool from a single-dimensional local edge site to a multi-dimensional network of distributed edge sites. By organizing edge resources across multiple geographic locations into a unified virtualized pool, the system maintains the proximity advantage of distributed edge computing while overcoming the resource limitation of individual sites through cross-site resource sharing and allocation.
Data Source
AI summary
A method includes receiving inputs for respective users in an edge utility system comprising edge and core computing sites, with a first one of the inputs for a first user characterizing edge resources requested by that user for executing at least a portion of a workload of that user, and a second one of the inputs for a second user characterizing edge resources available from that user for executing at least a portion of a workload of another user. The method includes populating one or more data structures based at least in part on the received inputs, aggregating edge resources of multiple ones of the edge computing sites into an edge network based at least in part on the populated data structures, and utilizing at least a portion of the aggregated edge resources of the edge network to execute at least a portion of a workload of a particular user.