Location tracking on mobile devices like smartphones has already begun to revolutionize personal navigation. Unfortunately, these services perform poorly indoors when GPS signals are no longer available. Highly accurate indoor location tracking would enhance a wide variety of applications including: building navigation (malls, factories, airports), augmented reality, location-aware pervasive computing, targeted advertising, social networking, participatory sensing and could even support next generation beam forming MIMO wireless networks. Current indoor localization systems for smartphones often use RF signal strength from WiFi access points or Bluetooth Low Energy (BLE) beacons to fingerprint indoor locations. Such systems are sensitive to environmental changes and obstructions, require extensive training procedures and are limited in both absolute as well as semantic localization accuracy. We propose using audio signals in the ultrasound spectrum, just above the human hearing range, to provide ranging and localization for many off-the-shelf mobile devices that are equipped with microphones. Ultrasonic ranging provides several advantages over RF-based ranging and fingerprinting approaches, which make it attractive for indoor localization. A relatively low propagation speed and carrier frequency allow for precise propagation time measurements in software using commodity hardware. Acoustic signals also have a low penetration depth, which confines them to target areas for accurate semantic localization. In this dissertation we address several challenges related to acoustic localization, including system scalability, ranging and localization accuracy, energy efficiency, robustness to noise, elimination of human perceivable audio artifacts, efficient use of limited acoustic bandwidth and rapid deployment strategies.
| Identifer | oai:union.ndltd.org:cmu.edu/oai:repository.cmu.edu:dissertations-2053 |
| Date | 01 August 2017 |
| Creators | Lazik, Patrick J.E. |
| Publisher | Research Showcase @ CMU |
| Source Sets | Carnegie Mellon University |
| Detected Language | English |
| Type | text |
| Format | application/pdf |
| Source | Dissertations |
Page generated in 0.0019 seconds