Research On Pedestrian Positioning And Navigation Technology Based On Smartphones In Indoor Environments

In the age of information technology, the ability to position and navigate persons in indoor environments has become increasingly important for a rising number of applications. Global Positioning Systems performs excellent in outdoor situations, but more and more market applications require seamless positioning capabilities in all environments. Therefore, research and development on indoor positioning has become a focus concern of different groups during the past decade.It has by now become apparent that several pedestrian indoor positioning technologies have been provided. Among these techniques, this thesis concerns on Pedestrian Dead Reckoning (PDR). One of the most important issues in PDR systems is heading estimation. Heading estimation is error-prone for indoor PDR systems. On one side, heading provided by magnetometers is in general not reliable for distortion of magnetic fields by mental and magnetic objects. On the other side, bias drift and other errors will be accumulated through integration, degrading orientation estimation accuracy when gyroscope is applied.For PDR systems with smartphones in pockets, one new factor deteriorates heading estimation accuracy that mobile phones will swing a little during people movement. Converting gyroscope readings into the world axis systems will not help, for unreliable readings of magnetometers must be included for conversion.Concentrating on trousers front pockets where mobile phones are usually carried, an interesting observation is found that swing of mobile phones will cause readings of gyroscopes lower than real angular velocity of person movement during turning or moving on curved paths. We call this fact Turn Insufficient Effect (TIE) and prove it with geometric analysis.Taking TIE into account, this thesis presents a Pedestrian navigation system using one Smartphone residing in trouser Pocket (PSP), only exploiting built-in accelerometers and gyroscopes. PSP extends the feedback control system of HDE (Heuristic Drift Elimination), compensating not only for drift bias of gyroscopes but also for Turn Insufficient Effect. A prototype of PSP is deployed on an Android based smart phone and evaluated with five participants under two indoor trajectories. The experimental results show that the average returning position error of PSP is1.33%and81%lower than HDE approach on the trajectory with repeated circles.