Aiming at the shortcomings of the existing indoor location algorithm, such as low accuracy of positioning, high deployment and maintenance cost, and unstable robustness, this paper proposes a method of indoor location...Aiming at the shortcomings of the existing indoor location algorithm, such as low accuracy of positioning, high deployment and maintenance cost, and unstable robustness, this paper proposes a method of indoor location based on the integration of smartphone with WiFi and magnetic field using multi-sensor fusion. In the initial stages of positioning, rough location is achieved by Wi-Fi-RSSI fingerprints which provides an initial location and geomagnetic matching area for indoor positioning based on particle filter magnetic field matching. This paper proposes the use of median filter algorithm to deal with the original magnetic field data and covariance interpolation algorithm to generate magnetic field map, and effectively reduce the interference which caused by geomagnetic fluctuations, thereby it will improves the positioning accuracy. Finally, through conducting comprehensive experiments and tests, the results show that the proposed technique can reliably achieve 0.836 meters precision in current experimental environment.展开更多
文摘Aiming at the shortcomings of the existing indoor location algorithm, such as low accuracy of positioning, high deployment and maintenance cost, and unstable robustness, this paper proposes a method of indoor location based on the integration of smartphone with WiFi and magnetic field using multi-sensor fusion. In the initial stages of positioning, rough location is achieved by Wi-Fi-RSSI fingerprints which provides an initial location and geomagnetic matching area for indoor positioning based on particle filter magnetic field matching. This paper proposes the use of median filter algorithm to deal with the original magnetic field data and covariance interpolation algorithm to generate magnetic field map, and effectively reduce the interference which caused by geomagnetic fluctuations, thereby it will improves the positioning accuracy. Finally, through conducting comprehensive experiments and tests, the results show that the proposed technique can reliably achieve 0.836 meters precision in current experimental environment.
