Design And Implementation Of The Identification Information Extraction System Of Internet Of Things Devices

With the rapid development of the Internet of Things technology,the security issues of the Internet of Things have become increasingly prominent.One of the most important problems is the exposure of Internet of Things assets.The Internet of Things devices exposed on the public network can easily be exploited by attackers once they have security vulnerabilities,causing serious consequences.Research has found that IoT devices of the same vendor,series,or model often have the same security vulnerabilities.Therefore,fine-grained identification of information such as vendor,type,and model of Internet of Things devices in cyberspace is essential to ensure the security of the Internet of Things.This paper designs and implements a rule-based IoT device identification information extraction system.Compared with the existing research on fine-grained identification of IoT devices,this system has improved the recognition effect.The system starts from the HTTP and HTTPS response data of IoT devices,combined with the assistance of search engines,performs fine-grained device identification,and extracts device vendor,type,and model information.The specific work of this paper mainly includes:First of all,this paper analyzes the existing research on the identification of Internet of Things devices,summarizes the shortcomings and deficiencies of the existing research on the identification effect,and puts forward some of the optimization points of this paper.Secondly,this paper designs and implements a rule-based IoT device identification information extraction system.Its core modules include:1.Data preprocessing module.The module first filters out the HTTP(S)responses of non-IoT devices,then integrates multi-dimensional features to extract text that may contain device information from the remaining HTTP(S)responses,and finally filters out irrelevant strings in the text.2.Candidate device keyword extraction module.This module extracts candidate device vendor,type,and model keywords from the text obtained by the data preprocessing module based on the rule library.In the extraction process,the device information contained in the special device fingerprint is considered,and the efficient extraction of O(N)time complexity is realized based on the FlashText algorithm.3.Device information extraction module based on search engine.The candidate device keywords extracted from the HTTP(S)response alone may not be able to determine the final device information.The module uses a search engine to assist in determining the final device information.Existing research has shortcomings in search strategies,methods of extracting device information from search results,etc.This paper has improved and innovated existing methods,and proposed a new set of methods that combine search engines to assist in determining device information,achieving a better recognition effect.4.Redirection and dynamic HTML response processing module.A considerable part of the HTTP(S)responses of IoT devices has redirection and dynamic HTML phenomena.This paper has conducted a comprehensive analysis and processing of these two phenomena.Compared with the existing research,more IoT devices can be identified.Finally,design the function and performance verification experiment for the identification information extraction system of the Internet of Things devices in this paper.It verifies the feasibility and effectiveness of the device identification method in this paper,and also verifies that the device identification method in this paper can achieve a better recognition effect than the existing methods.