Indoor Positioning System Design And Research Based On WLAN Fingerprint And Inertial Measurement

With the advances in wireless communication technology and socialdevelopment, the demand of wireless location-based service is growing. Satellitenavigation is a traditional way of positioning, it can usually provide solutions in theoutdoor scene, but in the scene where the building shelter is very serious, it does notwork any more, as a result, indoor positioning has become a hot research recentlyand the technology based on fingerprint of WLAN matures.Based on the exsiting technology, this paper proposed a new design method ofindoor positioning system.The designed system combined the results of WLANfingerprint positioning and inertial measurement results, two modes includingterminal independent positioning and server positioning mode were used infingerprint positioning, a method based on the euclidean distance of the accelerationvector was proposed to determine the current state of pedestrians, so that a meanfilter was used to calculate the result of the fingerprint positioning if the pedestrianwas in a stationary state.Then, an algorithm to detect gait which was based onacceleration peak detection was proposed and a step length estimation algotithm wasalso proposed to get the state of pedestrian. In terms of combining fingerprintpositioning and inertial measurement, an algorithm based on extended kalman fiterwas proposed. In terms of implementation of positioning system, this paperpresented the overall design architecture of the system, and the positioning softwaredesign based on Android mobile terminal was described in detail.Finally, the performance of the system was tested based on the system designand implementation, the results showed that the positioning accuracy got0.2metersbetter in the cumulative probability of1when the inertial measurement was usedin the stationary case. When the pedestrian was in moving state, the pedestriantracks got from the extended kalman fiter algorithm was closer to the true walkingtrajectory than that of the fingerprint positioning result in which inertialmeasurement was not used.The test verigied the feasibility and advantages of thedesigned positioning system.