PiPDR:A Placement-Independent Pedestrian Dead Reckoning System in Mobile Phones
Junzhao Du, Wen He, Jian Shen and Hui Liu
Usually, people source GPS for an accurate location. But, because the GPS signal is easily corrupted in urban canyons and indoor environments, locating a mobile user anytime and anywhere is still a demanding task. So, some researchers have tried to integrate some off-the-shelf inertial sensors in a mobile phone, like GPS, accelerometer and compass, to obtain a positioning solution by means of the Pedestrian Dead Reckoning(PDR) algorithm. However, most work implicitly required users to keep the phone’s placement fixed in hands in order to remove noise readings from phone sensors. In our paper, we design a novel placement-independent algorithm, called the PiPDR, to efficiently construct the PDR with a mobile phone. Except for the placement, it is also a challenge for us to track users without any a priori, user-specific knowledge such as the user’s walking direction and initial location. Our system solves the challenge by using a few scattered GPS positions or “encounters” with other phones. We implement the system on the Android-based platform and evaluate it comprehensively with data collected from more than 1050 traces.
Keywords: Pedestrian dead reckoning, mobile phone, placement-independent