WIFI indoor positioning method based on hybrid filtering and power exponent mapping

Abstract

The invention discloses a WIFI indoor positioning method based on hybrid filtering and power exponent mapping. The method comprises the following steps: firstly, performing Kalman filtering on received signal strength RSSI measured at an offline fingerprint collection stage, then rounding off the result to obtain an integer, performing Gaussian filtering to remove RSSI values with relatively small occurrence probability, then averaging the residual RSSI values, and finally figuring out -7 squares of all data to map the data into a power exponent domain. A K nearest neighbor algorithm is used as a positioning matching algorithm, thereby improving the positioning accuracy, and the calculated amount during the positioning is almost equal to that of the K nearest neighbor method, thereby being adaptive to positioning terminals with low computing performance and complex indoor electromagnetic environments.

Description

technical field [0001] The invention relates to the technical field of indoor positioning and navigation, in particular to a WIFI indoor positioning method based on hybrid filtering and power index mapping. [0002] technical background [0003] In recent years, with the gradual rise and popularity of smart portable mobile devices (such as smart phones, smart bracelets, etc.), location-based services and applications have developed rapidly and gradually expanded to all aspects of life travel. For example, electronic maps based on global positioning systems (such as GPS in the United States, Beidou navigation in China, etc.) provide people with positioning and navigation services in outdoor environments. Applications based on location information services emerge in endlessly (such as Baidu Maps, Gaode Maps, etc.). However, most of human activities are carried out indoors. Since the satellite signals of the GPS are difficult to penetrate buildings, the GPS cannot be used for i...

AI Technical Summary

Problems solved by technology

[0008] Finally, the value of RSSI is related to the detection device. The wireless signal transmitted by the same WIFI access point is detected by different receiving devices at the same point. The value of RSSI may be different. Therefore, the absolute value of the RSSI distance cannot be used only , to determine the weight value in the KNN algorithm

[0009] The "Science and Technology Bulletin", Volume 33, Issue 3, published in March 2017, introduced a method called "Indoor Positioning Method Based on Cosine Similarity Fingerprint Matching Algorithm", which does not use the Euclidean distance of the RSSI vector as a matching method, and the cosine value of the included angle of the RSSI vector (closest to 1) is used as the position matching method. Although this method is less computationally intensive than the KNN method to a certain extent, the positioning in some common and complex indoor office environments The accuracy is low, it is difficult to meet the positioning requirements

Images