Research On Indoor Positioning Technology By Low Power Bluetooth Based On Location Fingerprint With Filtering Noise

With the continuous development of mobile Internet technology and the widespread popularity of intelligent terminals, the demand for location-based services is becoming more and more intense. Whether indoors or outdoors, the basis of the location service is how to obtain the location information of the user, so quickly and accurately obtained location information and providing low cost location services become the direction of positioning technology development. Especially in recent years, people’s demand on indoor location services is intenser than the outdoor location services. stations, shopping malls and other public places also hope that through accurate indoor positioning to achieve the services based on the indoor positioning. But the current indoor positioning technology has the disadvantages of low positioning accuracy, high cost of deployment, so it is difficult to satisfy the large-scale application. Fingerprint positioning technology and low-power Bluetooth technology has advantages of simpler deployment, low cost and improving the accuracy of indoor positioning. Thus become a research hotspot.Indoor positioning means to implement the position location in the indoor environment, mainly use the base station location and wireless communication technology formed a set of indoor positioning system to determine the location of the indoor personnel. This paper aiming at the GPS positioning technology in indoor unable to get accurate location information, put forward a scheme of based on low power Bluetooth and the technology platform of fingerprint technology in indoor positioning, and for the shortcomings in the present stage of fingerprint positioning technology, we made some optimization on the offline training and the positioning stage.Based on the low power Bluetooth, we introduced the filter technology and k- nearest neighbor algorithm in fingerprint location, the main contents include:(1) Compare and analyse the technical characteristics of traditional Bluetooth and Bluetooth low energy, studied the fingerprint location algorithm that for commonly used, analysed the key step of the offline phase and positioning stage in the localization process.(2) Because this subject is in low-power Bluetooth platform to obtain signal strength parameters of Bluetooth devices to create a signal sample database. Therefore, we firstly analysed the distribution characteristics of the Bluetooth 4.0 signal, and the factors that impact the Bluetooth RSSI distribution, established the horizontal positioning error model that reflect the influence of the position error factors in indoor localization accuracy. According to the experimental results analysis, select the appropriate experimental environment for the improved scheme.(3) By analyzing the influence of the positioning error factors to the positioning accuracy, we propose using the Wavelet denoising, Kalman filtering method in the traditional fingerprint location process to weaken or eliminate noise signal caused by human activities. The experimental results show that the filtering algorithm has better filtering effect to the noise in the signal. Using Kalman filtering algorithm, the maximum error between the predicted point and the actual position is 1.1m. And the introduction of the wavelet denoising technology, the system for the static target positioning accuracy compared with the Calman filter algorithm is significantly improved, the maximum error is 0.7m.(4) Analysis of the traditional standard k- nearest neighbor defects, we propose a defined region k- neighbor algorithm design ideas, according to the proximity to divide the space and limite search area.