Employing Acoustic Properties to Authenticate Mobile Devices

The goal of this project is to use hardware components' built-in manufacturing faults as mobile phone IDs. We assessed the applicability of several I/O-related cell phone components, including sensors. Through this process, the focus was on creating hardware issue samples that could then be categorised using the device's speaker and microphone. In our technique, an audio sample was created by playing a known audio file via a mobile phone's speakers and then recording the sound using the same device. The impact of important variables on sample accuracy was examined using a variety of different sample groups. After collecting the samples, the frequency responses were extracted and classified. Data were categorised using a variety of classifiers, with certain label and sample group configurations achieving an accuracy of over 94.4%. The conclusions of this article suggest that speaker and mike production faults may be exploited for device authentication.