Multimodal Device-to-Device Ranging and Joint Localization

Location-based services (LBSs) have become integral to various smartphone applications, including turn-by-turn navigation, screen mirroring, and multidevice interaction. Nevertheless, traditional global positioning system (GPS)-based technologies face challenges in indoor environments, and the majority of current single-modality technologies suffer from various deficiencies, such as high deployment cost, short sensing distance, or long runtime latency. In this work, we present Bluetooth, acoustic sensors, inertial measurement unit (BAI), a smartphone-based multimodal ranging and localization system. To unleash the potential of multiple sensory signals, we design an enhanced Kalman filter (KF) algorithm to effectively fuse Bluetooth, ultrasounds, and inertial sensor signals for long-range measurement. Furthermore, as for multiple mobile devices in space with known relative distances, BAI constructs a network topology with mass-spring optimization to achieve joint localization. We implemented BAI on commercial smartphones and conducted extensive experiments in real-world scenarios to validate its effectiveness, robustness, and deployability. It achieves a mean absolute error (MAE) of 11 cm for ranging within 10 m and 12.5 cm for localization in a 6 × 6 m area. The demo video of BAI is available on YouTube (https://youtu.be/7Fbmn4ALaI0).