Abstract: In order to solve the problems that computing resources and wireless resources of terminal devices in the industrial Internet of things are difficult to meet the quality of service requirements and generate high energy consumption， a distributed deep learning-based resource allocation (DDLRA) algorithm is proposed. Firstly， the optimization problem of joint offloading decision and resource allocation of industrial Internet of things equipment is constructed. Then， the multiple parallel deep neural networks (DNNs) are used to solve the offloading decision and wireless resource allocation. Finally， simulation results show that the proposed DDLRA algorithm can improve task calculation speed and reduce energy consumption of terminal devices compared to the comparative algorithm.

Key words: industrial Internet of things； resource allocation； deep learning； mobile edge computing； energy consumption