A Booting Fingerprint of Device for Network Access Control

The rapid growth of the Internet of Things (IoT) is enhancing customers' experience with an ever-increasing mass of intelligent devices. However, the security doesn't get enough attention compared with novel concepts and features by manufactures which even not provide service of updates and patches. These issues could have been mitigated by the mechanism of the network access control. But traditional authentication mechanisms based on computational complexity of cryptographic protocols cannot be implemented on IoT devices with limited physical and computational resources. Therefore, we turn our attention to the device fingerprint which is like a biometric fingerprint that be able to provide identity authentication. The fingerprint proposed in this paper generated by network traffic features divided into the packet-feature and the IAT-feature. These features are extracted from the traffic when the device is booted, because at this time the device's network behavior is stable compared to other times. Experiments illustrate that the average identification accuracy of our device fingerprint reaches 92.4%.