Research On Device Independence In Indoor Localization Based On RSSI Technology

With the coverage of Wireless Local Area Network(WLAN)and the popularization of 4G network,people can access the Internet at anytime,anywhere,and get the information they need.Many applications based on location information also came into being,a new problem has cropped up.Owing to different mobile terminal device and implementation differences,the Received Signal Strength Indication(RSSI)data of Access Points(APs)collected by the mobile terminal device at the same location are different,thus cause the positioning error.First of all,collecting the WiFi data in the actual indoor environment and analyzing the differences between mobile terminal devices.In the process of analyzing,it is found that the abnormal RSSI values collected by mobile terminal devices is related to the instant environmental changes.Based on the above finding,the confidence interval is used to correct the abnormal RSSI value from the same wireless Access Point(AP)in this thesis and to improve the positioning accuracy.Secondly,exploring the solution to the problem of terminal device diversity.By comparing the RSSI data collected by different devices in different time periods,it was found that the RSSI values trend of different devices are similar when the different devices collect the AP wireless signal data in the same position.Based on the above findings,combined with Pearson correlation coefficient and the concept of weighting factor in Weighted K-nearest Neighbor(WKNN)algorithm,a method based on Pearson similarity is proposed.In the algorithm,calculating the Pearson correlation between the RSSI data collected by the positioning device and the fingerprint data of each sampling point in the location fingerprint database.The calculated results are used as a coefficient factor to solve the mismatch of the RSSI data collected by different devices.Finally,the effectiveness of the proposed Pearson similarity based device diversity elimination method has been verified based on the Android platform.The experiment uses three different mobile terminal devices to compare different existing algorithms.The results show that the proposed method can effectively reduce the divece diversity and indoor positioning errors,and improve the indoor positioning accuracy.The experimental results show that using the Pearson similarity based device diversity elimination method,the localization accuracy within 1.5m is more than 85% for homogeneous devices,and the positioning accuracy is maintained above 70% for heterogeneous devices.