Research On Integrated Indoor Localization Method By Dead Reckoning And Wi-Fi Based On The Smart Phone

In recent years,with the rapid development of the communication and the Internet technology,location-based services(LBS) have many very important applications in a lot of fields.The key technology of LBS is how to obtain location information. In the outdoor environment, GPS(Global Positioning System) can continuously provide high-precision location information in real time,and has become the main positioning technology outdoors.However,in the indoor environment, GPS navigation technology can’t provide high precision positioning because the signal is blocked. So an efficient and low-cost indoor localization method has become the focus of indoor positioning research. Most of existing indoor positioning methods need to decorate a lot of external devices, the cost of these methods are too high to realize. Using smart phones as a starting point, this paper put forward a combining Dead-Reckoning(DR),integrate step lengths and orientation estimations at each detected step to compute the absolute position and orientation of person, based on inertial measurement unit built-in smart phones and point scene Wi-Fi location method,and focuses on the following aspects.(1)Introducing the detection algorithm of pedestrian gait cycle. Firstly,pedestrians walking characteristics are analyzed,then comparing and analyzing the data measured by the tri-axial acceleration sensors embedded in smart phones and walk characteristics.Lastly,this paper expounds the gait detection algorithm based on local variance analysis.(2)Stride length calculation method based on velocity trace.Traditional stride length calculation method used the linear or nonlinear model based on gait frequency,which generated error quickly accumulated over time.To solve this problem,this paper designed stride length calculation method based on velocity trace.This method has an effective response to the step change of randomness and reduce the cumulative error with calculating the pedestrian walking speed and displacement in real-time.(3)Direction judgement based on direction sensor and gyroscopes.In the indoor environment,direction sensor will be affected by electromagnetic interference,and gyroscope can produce drift error because of itself defects, temperature change and other reasons. It will affect localization accuracy when sensor isexclusivly used.Therefore,this paper put forward a fusion method of the direction judgment based on gyroscope and direction sensor.(4)Wi-Fi correction methods based on scene choosing.Considering the complexity of the indoor environment and the characteristics of the Wi-Fi positioning method,we put forward Wi-Fi correction methods based on scene choosing.Adopting RSSI distance-measuring method in the long and narrow corridor and using the fusion method based on position fingerprint matching and RSSI distance-measuring in an empty room. It improved the positioning accuracy compared to that of simple positioning way.The results of the field measurement show that, the positioning accuracy and practicability of the combining the navigation calculation of inertial measurement unit built-in smart phones and points scene Wi-Fi location can better meet the requirement of indoor positioning.