High-accuracy indoor positioning systems provide various location-aware applications, enhancing our daily experiences. The Global Navigation Satellite System is difficult to use indoors, leading to the development of various indoor positioning methods. Among these, acoustic fingerprint-based positioning, which utilizes widely available speakers and microphones, provides robust performance in Non-Line-of-Sight (NLOS) environments that are common due to structural obstacles and furniture. These approaches typically use Received Signal Strength Indicator or Power Spectral Density as location fingerprints but face challenges in achieving high positioning accuracy. In this paper, we propose Geometric Sound Profile (GSP), a temporal feature of complex reflection waves, enabling high-accuracy localization with a single speaker. GSP, defined as the envelope of cross-correlation between the transmitted and received signals, starting from the transmission time, serves as a highly informative feature encapsulating the multipath-Time-of-Flight. Additionally, we implement a Convolutional Neural Network for the estimation of the user’s position using GSP. We generated a high-resolution pre-trained model in the simulation and fine-tuned it with measurement data, allowing accurate positioning with minimal measured training data. Our experiments demonstrated that the median positioning error was 0.14 m and the 90th percentile error was 1.23 m in the Line-of-Sight environment, and the median positioning error was 0.09 m and the 90th percentile error was 1.14 m in the NLOS environment.

Low-cost and accurate indoor location information can add spatiotemporal context to information systems, enabling new location-aware applications. Time-of-Flight (ToF)-based acoustic localization using speakers and microphones allows for localization accuracy within a few tens of centimeters, outperforming RF-based techniques. However, ToF-based localization requires synchronization between the speaker and microphone, i.e., the time offset between them must be known. Previous time offset estimation methods required custom hardware for speakers, limiting their practical use. Estimating the time offset using a single, unmodified speaker is essential for leveraging widely deployed speakers and enhancing coverage. This paper presents the first method for time offset estimation using a single speaker and a microphone, enabled by two key factors: (i) a time offset computation method that utilizes higher-order floor-ceiling reflections as multiple geometrically-constrained virtual speakers, and (ii) a signal processing pipeline that isolates these critical reflections from numerous others by leveraging the speaker's frequency-dependent radiation pattern. Experiments show that the proposed technique can achieve time offset estimation with a 90th percentile error of 259 μs at a 5 m distance. Furthermore, we implemented a ToF localization system based on SyncEcho, demonstrating a 11.0 cm localization accuracy with a 90th percentile error.

Robust indoor positioning systems provide stable location-aware applications, enhancing our daily experiences. Fingerprint-based positioning techniques enable estimation of a user's position in complex indoor environments. While previous studies have used the received signal strength indicator or power spectral density as fingerprints, they typically achieved only submeter accuracy. This paper presents Geometric Sound Profile (GSP) as a novel location fingerprint to elevate the performance ceiling of fingerprint-based positioning. GSP is derived from the cross-correlation of transmitted and received signals based on transmission time, and a user's position is computed using weighted k-nearest neighbors. Our experiments demonstrate a median error of 0.66 m, marking a significant advancement over previous fingerprinting techniques.