Edge-Cloud SNN-CNN Task Processing for Accuracy Bottlenecks
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Spiking neural networks (SNNs) are limited by low accuracy and complexity in processing tasks, while convolutional neural networks (CNNs) face high power consumption and hardware requirements, failing to provide an ideal user experience.
Innovation Solution
A hybrid approach combining SNNs and CNNs, where SNNs are deployed at edge computing nodes for low power consumption and fast processing, with CNNs at cloud nodes for high accuracy, enabling a feedback loop for improving SNN performance over time.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Use of energy by moving object
If spiking neural networks are used for task processing, then power consumption is reduced and hardware requirements are lowered, but processing accuracy and task complexity are insufficient
Solution Approach 1:
The system segments task processing into two parts: edge computing nodes use spiking neural networks for simple tasks to save power, while cloud computing nodes use convolutional neural networks for complex tasks requiring high accuracy. This segmentation allows each component to operate in its optimal performance zone.
Solution Approach 2:
The patent introduces a hierarchical architecture dimension, transitioning from single-node processing to a multi-level system spanning edge and cloud computing. This dimensional expansion enables the system to leverage both low-power edge devices and high-power cloud resources according to task requirements.
2Measurement precision
If convolutional neural networks are used for task processing, then processing accuracy and task complexity are improved, but power consumption and hardware requirements increase
Solution Approach 1:
The system segments computational workload based on task complexity and accuracy requirements, deploying convolutional neural networks only at cloud computing nodes where high accuracy is needed, while using simpler spiking neural networks at edge nodes for power-efficient processing.
Solution Approach 2:
The patent introduces an intermediary mechanism where edge computing nodes first attempt to process tasks using spiking neural networks, and only forward to cloud computing nodes when higher accuracy is required. This intermediary filtering reduces unnecessary high-power processing.
3Productivity
If spiking neural networks are deployed at edge computing nodes, then low power consumption and fast processing are achieved, but accuracy is insufficient for complex tasks
Solution Approach 1:
The system dynamically adjusts the processing architecture based on task characteristics. Simple tasks are handled by spiking neural networks at edge nodes for fast processing, while complex tasks are offloaded to cloud nodes with convolutional neural networks for high accuracy, creating a dynamic workload distribution system.
Data Source
AI summary
Embodiments of the present disclosure relate to a task processing method, a task processing device, and a computer program product. The method includes: receiving a task to be processed by a first neural network at a first computing node; determining that the first neural network failed to process the task; and sending the task to a second computing node for processing by a second neural network at the second computing node, wherein compared with the first neural network, the second neural network has at least one of the following: higher processing consumption and higher processing performance. With the technical solution of the present disclosure, a task can be processed quickly and accurately using a neural network with low processing consumption.