Print Email Facebook Twitter Mitigating IoT data management security concerns through blockchain and machine learning based solutions Title Mitigating IoT data management security concerns through blockchain and machine learning based solutions: Study and Conceptual Design Author van den Eeden, Lars (TU Delft Electrical Engineering, Mathematics and Computer Science) Contributor Lal, C. (mentor) Conti, M. (graduation committee) Degree granting institution Delft University of Technology Programme Computer Science and Engineering Project CSE3000 Research Project Date 2023-02-10 Abstract The Internet of Things industry is expanding rapidly. However, many security breaches occur, and privacy is often at stake in traditional IoT networks. These centralized systems will not be able to cope with dynamically changing environments. In light of these risks, it is crucial to prevent and minimize the chances of attacks occurring. Researchers have attempted to use blockchain for IoT security to ensure data consistency and availability. Fully public decentralized solutions for IoT still face data breaches. On the other hand, machine learning models detect potential attacks to create an effective defense system. This paper surveys state-of-the-art works looking to integrate blockchain with machine learning to protect data management for the IoT. Before exploring the various implementations, an analysis of multiple surveys that dive deeper into such integrations is made; then, five different blockchain and machine learning integrations. Many papers need a complete security analysis, and the experiments are limited. From studying the relevant integrations, this article presents a new scheme to protect IoT data using the DQNSB consensus algorithm to train a global model by distributing machine learning tasks, leveraging transparency to guarantee security. Subject BlockchainMachine LearningSecurityInternet of ThingsData ManagementScalability To reference this document use: http://resolver.tudelft.nl/uuid:2253c05b-ae0d-4a66-871d-ecd0b62ba1cf Part of collection Student theses Document type bachelor thesis Rights © 2023 Lars van den Eeden Files PDF final_paper_lars_van_den_eeden.pdf 1.72 MB Close viewer /islandora/object/uuid:2253c05b-ae0d-4a66-871d-ecd0b62ba1cf/datastream/OBJ/view