Research On Protocol Level Energy Consumption Of Smart Home Edge Gateway

With the advent of the "Industry 4.0" era,all walks of life have used Internet of Things,big data,cloud computing and other technologies to reform,and people's production and life are gradually becoming intelligent.Subsequently,because of the complexity of the application,the performance requirements of embedded devices have been gradually improved and the number of embedded devices has also increased dramatically.The huge number of these devices will consume a large amount of electricity,which will lead to more and more prominent problems such as climate warming and environmental pollution.Nowadays,the world is vigorously promoting "low carbon emission reduction".Therefore,this paper studies the energy consumption at the protocol level of the multi-protocol edge gateway that will be widely used.In smart home industry,because of the lack of uniform specification,manufacturers may adopt different communication protocols,and the interconnection between devices is difficult.Therefore,gateways that can connect multiple heterogeneous devices have become one of the most important nodes.At the same time,due to the insufficiency of cloud computing technology for the processing of Internet of Things data,the edge computing technology is gradually developed,and the gateway will become an important edge node in the smart home system.This paper analyzes the energy consumption of the protocol level in the way that the main controller uses the software to parse the protocol and the hardware protocol stack chip to parse the protocol.Firstly,this paper introduces the research background and significance of energy consumption analysis,and analyzes the research on the energy consumption of embedded devices by domestic and foreign scholars.Secondly,the edge computing technology and its architecture and application are introduced in detail,and a new edge IoT four-tier architecture is proposed,including edge nodes,edge servers,cloud centers,and smart applications.Then,for the software protocol parsing method,the program is analyzed by using the symbol execution,the unreachable path is eliminated.And the source code is compiled into instruction set,and the energy consumption at the protocol level of smart home gateway is analyzed from three aspects: energy consumption of a single instruction,energy consumption between instructions,and other energy consumption to construct the energy consumption analysis model.The Modbus/TCP and Modbus/RTU protocols are mutually converted on the STM32F103 experimental board.The feasibility and accuracy of the model are tested by comparing the results of the three methods.Finally,according to the hardware protocol parsing method,a multi-protocol edge gateway based on hardware protocol chip is designed to realize the mutual conversion of Ethernet,Wi-Fi,ZigBee and CAN protocols.The energy consumption of the protocol level is calculated when the gateway converts different types of protocols and different number and size of data packets.In the paper,the BP neural network is used to train the energy consumption prediction model,and the model is verified and analyzed.The model proposed in this paper can analyze the energy consumption of the multi-protocol edge gateway software and hardware protocol in the smart home system in the edge IoT environment.The model has high accuracy and can provide guidance for the design of multi-protocol edge gateway with low power consumption.