Research Article
IoT Devices Classification Base on Network Behavior Analysis
@INPROCEEDINGS{10.1007/978-3-031-17081-2_11, author={Lingan Chen and Xiaobin Tan and Chuang Peng and Mingye Zhu and Zhenghuan Xu and Shuangwu Chen}, title={IoT Devices Classification Base on Network Behavior Analysis}, proceedings={Applied Cryptography in Computer and Communications. Second EAI International Conference, AC3 2022, Virtual Event, May 14-15, 2022, Proceedings}, proceedings_a={AC3}, year={2022}, month={10}, keywords={Internet of Thing Machine learning Deep learning Traffic classification}, doi={10.1007/978-3-031-17081-2_11} }- Lingan Chen
Xiaobin Tan
Chuang Peng
Mingye Zhu
Zhenghuan Xu
Shuangwu Chen
Year: 2022
IoT Devices Classification Base on Network Behavior Analysis
AC3
Springer
DOI: 10.1007/978-3-031-17081-2_11
Abstract
IoT devices classification from the network traffic have attract increasing attention due to the manage requirement of the growing IoT applications.
Current statistic feature based methods needs to sample most of the packets and requires much computation during the feature extracting. The features are easily affected by the network environment and user operations.
This paper proposes a network behavior analysis (NBA) based IoT devices classification scheme which takes the sequence of network session features as input and learns the behavior feature with LSTM for IoT devices classification. NBA bases on the network behavior analysis without parsing the packet of network traffic, thus can handle the traffic using encryption protocol. In addition, we built a testbed with IoT devices for evaluation.
The traffic generated in the testbed was captured for evaluation. The result of the experiment indicates that by incorporating destination and interval features, our method can accurately classify IoT and non-IoT devices and achieves the accuracy over 99%.
