Edge Device Latency Reduction via Spatial Network Orchestration
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional edge computing networks face high latencies and performance bottlenecks due to a single point of failure and lack of collaboration among edge devices, leading to inefficient task offloading and increased energy consumption, especially in dynamic environments like smart traffic signal control systems.
Innovation Solution
A spatial network community-based edge orchestration system using deep reinforcement learning (DRL) with online policy gradient training and importance sampling, along with a reactive distributed request for help method, to optimize task offloading and resource utilization across edge devices, reducing latency and energy consumption.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If conventional edge computing networks use a single centralized orchestrator for task offloading, then the system structure is simple, but latency increases and reliability decreases due to single point of failure
Solution Approach 1:
The patent segments the centralized orchestrator into multiple distributed edge devices, each capable of autonomous task offloading decisions. This segmentation eliminates the single point of failure while maintaining manageable complexity through local decision-making logic at each edge device.
Solution Approach 2:
Each edge device performs self-service by autonomously evaluating task requirements and making offloading decisions based on local conditions. This self-service capability eliminates dependency on a centralized orchestrator, improving reliability while keeping individual device complexity low through standardized autonomous decision protocols.
2Productivity
If edge devices operate independently without collaboration, then device complexity is low, but task offloading efficiency decreases and latency increases
Solution Approach 1:
The patent merges the capabilities of multiple independent edge devices into a coordinated collaborative system. Edge devices share task information and collaboration status, enabling efficient task offloading while maintaining relatively simple individual device architectures through standardized interaction protocols.
3Productivity
If edge devices continuously monitor and communicate with each other, then task offloading optimization improves, but energy consumption increases
Solution Approach 1:
The patent implements periodic status updates and monitoring instead of continuous communication. Edge devices exchange information at scheduled intervals, achieving sufficient optimization for task offloading while significantly reducing energy consumption compared to continuous monitoring and communication protocols.
Data Source
AI summary
Deep reinforcement learning is applied to self-orchestration in edge device computing for offloading within a spatial network community to reduce latency and bandwidth issues. A revised online policy gradient training algorithm based on importance sampling in addition to the use of DRL-based offloading provides for continued use of original sample training data. A request for help scheme supports edge-device cooperation among neighboring devices of the spatial network community by sharing edge device state information (EDSI) for governing task assignments.