A hybrid information-based smartphone indoor-position approach

Abstract

Indoor smartphone positioning is one of the key ICT techniques. Generally, the development and implementation of an indoor positioning system heavily relies on the technology of wireless sensor network (WSN) since wireless sensors can estimate the probable distance between radio source and the sensors themselves by evaluating the strengths of wireless signals, received from the radio source, such as RSSIs of Wi-Fi or Bluetooth. However, the wireless signals could be influenced by indoor and outdoor objects, and carrying mode of a user’s smartphone, like in-pocket and in-backpack. Therefore, this study proposes an indoor positioning scheme, named LEarning-based Indoor Positing System (LEIPS), which identifies the carrying mode of a user’s smartphone by using this smartphone’s inertial sensors, aiming to increase the positioning accuracy of indoor positioning. Deep learning algorithms are also deployed by the LEIPS to improve the prediction performance. The experimental results demonstrate that this system can reach 99% of positioning accuracy. In the meantime, carrying mode information is validated to show that it is able to improve the accuracy of a positioning system.