Year: 2024
Type: article
Abstract: Criminals exploit the robust anonymity afforded by Tor for illicit purposes, prompting heightened interest among researchers in de-anonymization attacks on the Tor network. The execution of experiments on de-anonymization attacks within a real Tor network presents considerable challenges, hence the necessity for a simulation environment. However, existing methods for simulating the Tor network are inadequate regarding realism, flexibility, and scalability, with some being prohibitively expensive. In this paper, we develop a lightweight and scalable Tor network simulation environment based on Kubernetes (K8s), employing Docker containers to simulate Tor relays. The results demonstrate that a network of up to a thousand Tor relays can be simulated using just four standard hosts. Furthermore, two de-anonymization attack experiments were conducted within this simulated environment, which exhibited high levels of realism and flexibility. Finally, a hybrid networking approach combining multi-granularity relays was explored to enhance further the balance between realism and cost in Tor network simulations.
Source: Security and Safety
Authors: Wentao Huang, Han Wu, Zeyu Li, Xiaqing Xie, Qingfeng Zhang, Xuebin Wang, Jiapeng Zhao, Jinqiao Shi
Language: English
Subfield: Artificial Intelligence
Field: Computer Science
Domain: Physical Sciences
Open Access status: diamond
Awards: G2407591565, G7261400518, G889106969
OpenAlex ID: W4405934377