PROTECTING PRIVACY AGAINST TRAFFIC ANALYSIS USING MACHINE LEARNING METHODS IN IoT NETWORKS
Project no. 29550 · Status: Kapanmış · 2023
Summary
<p>The role of technological devices in people's lives has gained importance with the development of technology. The necessity of devices that use the Internet has become inevitable with the increase in Internet usage and the development of technological devices. The Internet of Things (IoT) is a connected network where physical devices and systems embedded with electronic hardware, sensors and software communicate with each other and exchange data. These devices can be household items, personal devices or industrial tools. The data transfer between these devices can be achieved over the network environment due to the communication between the devices by accessing the Internet. In addition to non-sensitive information about users, this data may also contain extremely sensitive information such as home address and card password. Therefore, threats against the users' privacy occur. Even if IoT network traffic is encrypted, users' personal data can be obtained from IoT devices by analyzing network packets. In this project, it is aimed to achieve the balance between privacy and utility in terms of network performance in the context of IoT device management. For this purpose, Differential Privacy and Padding methods will be used to prevent IoT data from being captured and used by malicious individuals who analyze network traffic using Machine Learning methods and to protect user privacy.</p>