Analyzing the Feasibility and Generalizability of Fingerprinting Internet of Things Devices

Abstract

In recent years, we have seen rapid growth in the use and adoption of Internet of Things (IoT) devices. However, some loT devices are sensitive in nature, and simply knowing what devices a user owns can have security and privacy implications. Researchers have, therefore, looked at fingerprinting loT devices and their activities from encrypted network traffic. In this paper, we analyze the feasibility of fingerprinting IoT devices and evaluate the robustness of such fingerprinting approach across multiple independent datasets — collected under different settings. We show that not only is it possible to effectively fingerprint 188 loT devices (with over 97% accuracy), but also to do so even with multiple instances of the same make-and-model device. We also analyze the extent to which temporal, spatial and data-collection-methodology differences impact fingerprinting accuracy. Our analysis sheds light on features that are more robust against varying conditions. Lastly, we comprehensively analyze the performance of our approach under an open-world setting and propose ways in which an adversary can enhance their odds of inferring additional information about unseen devices (e.g., similar devices manufactured by the same company).

Identifying key sectors based on cascading effect along paths in the embodied CO2 emission flow network in Beijing-Tianjin-Hebei region, China

“…Considering the computing cost brought by a large number of features, works published in 2021 to 2022 [80,110,111] focus on feature reduction algorithms to acquire key features that enhance classification accuracy and efficiency. Moreover, they also considered greater device diversity and larger datasets [69,111,112].…”

“…Ahmed et al [69] unprecedentedly considered a remarkable number of 188 devices. The experiment integrated six public datasets along with a self-collected dataset (the "Ours" dataset).…”

“…Moreover, recent papers have specifically studied feature reduction techniques, which have also made breakthroughs in efficiency. In practical application scenarios, they still face a significant issue: traffic characteristics may be confused between devices of the same type but different models [69]. However, only a small part of the literature focuses on devices using low-energy protocols such as Bluetooth and Zigbee.…”

The Protection of Data Sharing for Privacy in Financial Vision

“…Many privacy-preserving methods can prevent the re-identification problem of a particular individual's private information caused by comparisons or queries with other databases. The most widely used one is differential privacy [14]. Many federated learning models are now used with differential privacy mechanisms to minimize the risk of data leakage [4,15].…”