Loading...

Privacy Preserving against Wireless Devices Sniffer in Smart Home

Kazem, Ameneh | 2025

0 Viewed
  1. Type of Document: M.Sc. Thesis
  2. Language: Farsi
  3. Document No: 58546 (19)
  4. University: Sharif University of Technology
  5. Department: Computer Engineering
  6. Advisor(s): Amini, Morteza
  7. Abstract:
  8. Smart homes, due to their extensive use of Internet of Things (IoT) devices and wireless communication protocols, are highly vulnerable to metadata leakage that can reveal users’ behavioral patterns. Although recent studies have proposed various countermeasures such as encryption or random traffic injection, these solutions often exhibit limited effectiveness due to temporal and logical inconsistencies between real and synthetic data or the need for modifications in device infrastructures. To address these challenges, this research introduces the {MAZE} tool as a non-intrusive and efficient framework that employs a two-level Continuous-Time Markov Chain (CTMC) modeling approach to simulate user behavior, generate realistic synthetic traffic, and inject it into the network while maintaining strong temporal and logical consistency with real traffic. The method involves extracting activity transition matrices, start-time distributions, and duration models from real datasets, generating sequences of activities and their internal events, converting them into Zigbee-compliant packets, and broadcasting them alongside real traffic. Simulation and field experiment results show that injecting {50\%} synthetic traffic can reduce the accuracy of transformer-based adversarial models from approximately {97.5\%} to around {20\%}, demonstrating the tool’s effectiveness in significantly degrading an attacker’s inference capabilities. Therefore, MAZ provides a practical, low-cost, and highly effective solution for preserving user privacy in smart homes without requiring any modifications to existing devices or communication protocols
  9. Keywords:
  10. Smart Home ; Internet of Things ; Privacy ; Zigbee Network ; Continuous Time Markov Chain ; Synthetic Traffic

 Digital Object List

 Bookmark

...see more