Research On Early Detection Method Of Node Capture Attack In Internet Of Things

At present,the Internet of Things is profoundly changing people's production and life,social management and public services,and is accelerating.The Internet of Things is also carrying an increasingly large amount of user data.Therefore,How to protect the network information security of the Internet of Things has always been the research focus of scholars.The small size,low power consumption,and characteristics of Io T node devices are often distributed in unattended corners,making them the main entrance for attackers to control the network.For devices at the perception layer of the Internet of Things,node capture attacks are one of the most vulnerable threats.In the past few decades,information security technology has developed rapidly,and more and more strategies to prevent node capture attacks have been continuously proposed or improved.From node location-based detection to malicious data injection detection,a large number of researchers have perfected the detection methods of node re-deployment back to the network stage in node capture attacks and malicious nodes launching internal attacks.In order to detect abnormal nodes as early as possible,this article will study the detection methods of node capture attacks in the early stage,and consider combining the characteristics of the most commonly used MQTT protocol in the Internet of Things with the core ideas of typical detection methods to achieve detection of Internet of Things perception The purpose of the layer abnormal node device.Firstly,this paper explains the research background and significance,and introduces the research status.Then it introduces the concept of node capture attack and the idea of typical detection methods.It also introduces the characteristics of the message type,network architecture,heartbeat mechanism and message service quality level of the MQTT protocol.In response to the problem of high false alarms in monitoring methods,this paper proposes an early detection method for node capture attacks under the server-client architecture.It is decided to use heartbeat messages and greeting messages to jointly monitor the status of nodes.The status is finally determined to avoid false alarms;in order to achieve a balance between detection rate and network traffic consumption,this paper also analyzes the relationship between the frequency of heartbeat messages and the total number of captured nodes in the network,and designs an adaptive heartbeat frequency Mechanism,and the effectiveness of the detection method and adaptive mechanism proposed in this paper is proved through experiments.In addition,this paper also applies the detection method to actual scenarios to verify the usability in real environments.Finally,this article summarizes the research done and prospects for future research.