Design And Implementation Of An Intelligent IoT Gateway Supporting Edge Computing

The Internet of Things has been widely used in smart cities as an important means to perceive the physical world.it is very common for us today to see fire in urban life.So building a comprehensive,safe,and effective firefighting Io T is of great significance to building a safe society.However,with the deepening development of the construction of smart cities,the scale of the Internet of Things for firefighting is getting bigger,the network topology of firefighting Io T is becoming more complex,and the variety of networked devices is becoming richer.Therefore,the current firefighting Io T based on cloud computing architecture is gradually unable to meet the needs of firefighting safety.According to the requirements of a city safety information solution provider,this thesis aims to design an intelligent Io T gateway with edge computing function for fire facility monitoring.This thesis uses the edge computing reference architecture proposed by the Edge Computing Consortium as a reference guide.At the same time,the edge computing architecture of intelligent firefighting Io T is designed based on the traditional firefighting Io T architecture system.In addition,considering the requirements of monitoring firefighting facility,the thesis designs a systemetic overall scheme consisting of intelligent Io T gateway,information collection nodes and information monitoring center.Besides,the hardware platform of the intelligent Io T gateway and the hardware platform of the information collection nodes are built,and the intelligent Io T gateway platform uses Jetson Nano as the platform controller equipped with Wi Fi communication module,Lo Ra communication module,4G communication module.At last,the information collection nodes are divided into two types of sensing nodes according to the nature of the collected data,using STM32F103Z8T6 and ESP32 as the core controllers of the terminal nodes,respectively.For software design,the thesis first uses Linux as the operating system to the intelligent Io T gateway platform,designing the data frame format for Lo Ra communication and UDP transmission to ensure the stability of data transmission between each node and the intelligent Io T gateway platform.In addition,data collection and transmission control procedures are designed for information collection nodes in 4 different types of environments.Besides,the information monitoring center realizes the visual display of data based on Web server and database.Finally,for the demand of detecting whether the firefighting passage is occupied or not,considering the impact of outdoor natural light variation on the accuracy of the detection algorithm,the low-light classifier is designed and the Zero-DCE low-light image enhancer is introduced.After combining the improved YOLOv4-tiny target detection algorithm,this thesis achieves the intelligent detection for outdoor firefighting passage occupancy.After being tested,all the design indicators of the system have been achieved,satisfying the functional requirements of the users.The intelligent Io T gateway not only realizes the data conversion between different transmission protocols,but the intelligent detection of fire passage occupied by vehicles.