Research And Implementation Of Data Security Filtering Component Based On MQTT Protocol

With the wide application of Internet of Things technology in power system,security incidents based on power Internet of Things occur from time to time,such as power core data leakage,power terminal equipment being hijacked and manipulated illegally,etc.At present,the security protection system in the field of power Internet of Things has not been perfected,and the Internet of Things management platform of power Internet of Things lacks the ability to identify edge Internet of Things agents,and has security risks such as illegal access of unauthorized edge Internet of Things agents and controlled by a fake master station.This paper conducts in-depth research on the MQTT protocol used in the power Internet of Things,analyzes the system architecture of the power Internet of Things and the existing security risks in detail,and designs a set of power Internet of Things security protection schemes based on the MQTT protocol for security risks.Thereby,the communication between the edge IoT agent of the power Internet of Things and the IoT management platform is securely strengthened.The main content of this paper is summarized as follows:(1)The MQTT protocol based on TLS is adopted to achieve secure communication,identity authentication is carried out when the edge IoT agent accesses the IoT management platform to prevent the access of illegal devices,and encrypted transmission of data is carried out after the negotiation,so as to realize the protection of privacy and integrity of data.(2)Through in-depth study of MQTT protocol,the deep packet detection module based on MQTT protocol is designed.The deep analysis is carried out for the three components of MQTT message:fixed header,variable header and payload,and this module can perform preliminary filtering on illegal messages(3)Aiming at the abnormal traffic in the power Internet of Things,the abnormal traffic detection module of the MQTT protocol based on the random forest algorithm is designed.The specific implemention methods are as follows:①Collect the communication data between the edge IoT agent and IoT management platform;Preprocess data and extract flow characteristics to obtain sample data;③ Use sample data to fully train and evaluate the classification model,and build a data model that meets the requirements of performance indicators;④ realize the dynamic monitoring of data between the IoT management platform and the edge IoT agent.The results show that each functional module can achieve the corresponding defense attack effect,which can effectively guarantee the security of the power Internet of Things.