Fingerprinting Mainstream Io T Platform Through Traffic Analysis

The Internet of Things(Io T)platform is an essential component in smart home application scenarios.Under the control of the Io T platform,Io T devices are connected with a remote cloud server,upload data,receive commands,and directly or indirectly(forwarded by remote cloud server)connect with the user's mobile terminal.While providing smart home solutions for Io T device manufacturers and smart home users,bringing unprecedented convenience,the existence of mainstream Io T platforms provides specific targets for malicious attackers.In view of this situation,we designed and implemented a fingerprinting method for mainstream Io T platforms to identify Io T entities inside computer systems.Specifically,the main contributions of this paper are:1.Investigated the design of architecture,business logic,and network security related mechanisms of several mainstream Io T platforms at home and abroad,especially the smart home platform,revealed the basic situation of these platforms,and proved the universality of security problems in Io T platforms.2.Introduced the concept of fingerprinting technology of the operating system,which is widely used in the traditional network security industry,into the research of Io T platforms for the first time,and analyzed the traffic of users' mobile terminals and remote servers in Io T platforms which are widely used manually.In this process,we designed a set of workflows that suitable for the current mainstream Io T platform smart home solutions of traffic capturing,feature extraction and fingerprint model construction,collected a large amount of traffic generated during the actual operation of smart home scenes,and extracted features from network traffic content to distinguish different Io T platform service providers,or to discriminate Io T traffic from non-Io T traffic as fingerprints,so as to build a fingerprinting feature model for the traffic between users' mobile terminals and servers of mainstream Io T platforms.3.Designed and implemented a software tool for marking the traffic between the mobile terminal and remote server of mainstream Io T platforms in network traffic,and tested the usability and performance of the marking tool through running marking program with actual pre-marked traffic.4.Made a conclusion on this research,discuss several significant application scenarios of fingerprinting technologies and the marking tools,and look forward towards the possible future work.