Device Identification Based On Power Waveform Using USB-PD Negotiation

The widespread adoption of USB Type-C technology has been remarkable in recent years. Electronic devices such as smartphones, tablets, and computers now utilize Universal Serial Bus (USB) Type-C connectors in compliance with the USB Power Delivery (USB-PD) specification for power transmission and reception. Although USB Type-C devices are equipped with a function of auto-notification of their IDs, the function is not always guaranteed to work; we cannot rely on the ID notification. Relying exclusively on this data for comprehensive device identification presents a significant obstacle. Based on the ground, this study introduces an approach to identify USB-C-connected devices by analyzing the power waveform in USB PD negotiations, a distinctive feature of USB Type-C. The goal is to identify these devices regardless of their charging rates or circuit designs. Furthermore, our proposed method was validated using a virtual grid hub (VG-Hub) with seven USB-PD-compliant TypeC ports oriented toward renewable energy. The proposed method using VG-Hub showed excellent performance, achieving 99.25% accuracy, repeatability, and F1 score over the previous method. This demonstrated the effectiveness of the USB Type-C connected device classification method independent of device state.