Collaborative deep learning reasoning method for decentralized equipment

Abstract

The invention discloses a method for deploying a cache-based deep neural network on decentralized edge equipment in a distributed manner. The method comprises the following steps: dividing a neural network; pruning is carried out on the previous layer of neural network at the division position; then, a part of the deep neural network is calculated in the task initiating device; a small number of intermediate results are transmitted to other edge equipment; and calculating the remaining portion, in addition, caching and reusing an intermediate result of the edge device neural network; differentdevices can share cache, so that the delay of an edge intelligent application is reduced, the requirement of a neural network for the performance of the edge device is reduced, and particularly, whenan intelligent task request is initiated to similar data on the edge side, the repeated calculation amount can be reduced, the performance requirement of deep learning for the device is reduced, andthe calculation resources of an edge scene are fully utilized.

Description

technical field [0001] The invention relates to the fields of artificial intelligence and edge computing in computer science, in particular to a deep learning reasoning method that combines collaborative computing and caching of distributed devices. Background technique [0002] As an emerging computing paradigm, edge computing aims to use the computing and communication resources of edge devices to meet the needs of users for real-time response to services, privacy and security, and computing autonomy. Driven by the rapid development of algorithms, computing power and big data, as the most active field in artificial intelligence, deep learning has made significant progress in many fields. With the development of the Internet of Things and cyber-physical systems (CPS), new applications such as autonomous driving, intelligent drone formation, and intelligent robot clusters drive the integration of edge computing and artificial intelligence, and promote the emergence and rapid...

AI Technical Summary

Problems solved by technology

[0007] These methods can accelerate the deep learning model in a certain way, but they cannot make full use of the characteristics of the interconnection between edge devices, nor do they take into account that the input data of edge devices that are likely to perform intelligent tasks within a period of time is relatively similar or even repeated.

Images