Charge-Depleting of the Batteries Makes Smartphones Recognizable
Jing Chen, Yingying Fang, Kun He, Ruiying Du · 2017
Many components of smartphones are used to generate device fingerprinting, such as screens, CPUs and various sensors. These device fingerprinting can be used to identify the smartphones. However, there are many restrictions with these device fingerprinting. Invariable information in screens and CPUs may lead to privacy risks. Moreover, strict experimental steps are required when fingerprinting the sensors. The effectiveness and effeciency of these device fingerprinting is reduced in practice. In this paper, we present a novel hardware fingerprinting based on the battery. Instead of relying on invariable information of the battery, we focus on the charge-depleting of the smartphone. The discrepencies on manufacturing of smartphones make that the charge-depleting is different when performs the same task. Moreover, charge-depleting information can easily be obtained without strict operating steps. We design a highly accurate algorithm to fingerprint the batteries which is based on the unsupervised learning. Besides, we stimulate the algorithm with different charge-depleting of tasks to improve the performance. We use 15 smartphones to evaluate the performance of the battery fingerprinting in both laboratory and public conditions. The experimental results show that battery fingerprinting is quite effective, the recognition accuracy rate can reach 86%.