Indoor Navigation on Smartphones
Vassilyev, Artem (2014)
2014
Master's Degree Programme in Information Technology
Tieto- ja sähkötekniikan tiedekunta - Faculty of Computing and Electrical Engineering
This publication is copyrighted. You may download, display and print it for Your own personal use. Commercial use is prohibited.
Hyväksymispäivämäärä
2014-11-05
Julkaisun pysyvä osoite on
https://urn.fi/URN:NBN:fi:tty-201410221512
https://urn.fi/URN:NBN:fi:tty-201410221512
Tiivistelmä
Today the main issue of absolute navigation system mainly consists of GNSS signal propagation problems in indoor areas and street canyons (also known as urban canyons). The existing solutions are unable to provide reliable location service in these areas thus a new approach is needed. Due to the rapid growth of smartphone market, wireless trans-mission networks (e.g. Wi-Fi, Bluetooth, WiMAX) have been gaining popularity over the last few years. These types of networks were originally designed for high-speed transmission of large data i.e. Internet access, but it can also be used for the navigation purposes. Moreover, during the evolution of smartphones, manufacturers started to add new types of self-contained sensors that have never been used in such a way before. Some of them like accelerometers, magnetometers and gyroscopes can be used to track movement and position of smartphone in space.
During this research one of the latest and the most sensor-equipped smartphone was tested. Nexus 5, released by Google, was utilized as a testing platform for indoor tracking application based on self-contained sensors only. This implies a highly laborious and tedious work of manually collected training data and developing the corresponding indoor tracking application. The methods used in development process allows decreasing the overall development costs while notably improving the performance of the existing navigation systems.
The implemented indoor navigation application utilizes pedestrian dead reckoning method that allows improving the accuracy of existing navigation methods. It can also be used separately in fingerprinting or SLAM process. This application was tested in several indoor areas with different location properties: narrow corridors, wide halls, tiny rooms. The corresponding application utilizes built-in accelerometers, magnetometers and step detectors to track the route. Magnetometer fluctuations were smoothed by using low-pass filter. The experiments showed the total positioning error between 7% and 14%, respectively. Tests of built-in step detector showed the average detection error of 0.5%, which is lower than existed solution can obtain. In general, the obtained positing error and performance improvement can be considered as immaterial but the results can be used as a platform for the future research.
During this research one of the latest and the most sensor-equipped smartphone was tested. Nexus 5, released by Google, was utilized as a testing platform for indoor tracking application based on self-contained sensors only. This implies a highly laborious and tedious work of manually collected training data and developing the corresponding indoor tracking application. The methods used in development process allows decreasing the overall development costs while notably improving the performance of the existing navigation systems.
The implemented indoor navigation application utilizes pedestrian dead reckoning method that allows improving the accuracy of existing navigation methods. It can also be used separately in fingerprinting or SLAM process. This application was tested in several indoor areas with different location properties: narrow corridors, wide halls, tiny rooms. The corresponding application utilizes built-in accelerometers, magnetometers and step detectors to track the route. Magnetometer fluctuations were smoothed by using low-pass filter. The experiments showed the total positioning error between 7% and 14%, respectively. Tests of built-in step detector showed the average detection error of 0.5%, which is lower than existed solution can obtain. In general, the obtained positing error and performance improvement can be considered as immaterial but the results can be used as a platform for the future research.