Research On Indoor Location Algorithm Based On WiFi And Bluetooth Fusion

With the popularization of mobile intelligent terminal devices and the rapid development of Internet of Things technology,location-aware technology plays an important role in people's lives and work.It has become a research hotspot in the field of location services and has broad application scenarios.Since the smartphone device supports WiFi and Bluetooth,WiFi and Bluetooth have become one of the most widely used indoor positioning technologies.WiFi has the advantages of low cost and wide coverage.Bluetooth has the advantages of low power consumption and high positioning accuracy.Therefore,the research on WiFi,Bluetooth and the fusion positioning technology of both has great theoretical and application value.In view of the problems that WiFi and Bluetooth signals are susceptible to interference during the propagation process and the positioning accuracy of the single positioning technology is limited,this thesis proposes a WiFi fingerprint positioning and Bluetooth fingerprint positioning algorithm,and on this basis,a WiFi and Bluetooth Fusion positioning algorithm is proposed.The content of this paper is as follows:(1)An AP weighted feature distance positioning algorithm based on WiFi fingerprint is designed.In the stage of offline fingerprint database construction,aiming at the problem of unstable WiFi signal,AP selection algorithm and mean filtering algorithm are used to select the optimal AP and preprocess AP.A fingerprint library construction method is adopted to store the standard deviation of the AP signal and the signal strength value as reference point fingerprint feature values into the fingerprint library.In the online algorithm matching stage,an AP-based weighted feature distance positioning algorithm is proposed.By calculating the AP's influence capability index,each AP is given a weight.Based on the distance between the reference points,the required reference points are screened and given weights.And calculate the position coordinates.Compared with the WKNN(Weighted K Nearest Neighbor)algorithm,the proposed algorithm has increased the probability within 3 m of positioning error from 79.6% to 87.5%.(2)A WKNN positioning algorithm based on Bluetooth fingerprint filtering is designed.In the construction stage of offline fingerprint library,for the problem of instability of the Bluetooth signal,the 3 criteria and Gaussian filter algorithm are used to preprocess the Bluetooth signal to remove the unstable and small probability of interfered signals,and then the Bluetooth signal and the reference point coordinates are used asfingerprint features.The value is stored in the fingerprint database,and the improved K-means clustering algorithm is used to divide the fingerprint database.In the online algorithm matching stage,a WKNN positioning algorithm based on fingerprint filtering is proposed.Based on the Bluetooth signals collected online,reference points that are not helpful for positioning accuracy are removed,and the position coordinates are calculated in conjunction with the WKNN algorithm.Compared with the WKNN algorithm,the proposed algorithm has increased the probability within 2 m of positioning error from63.4% to 76.6%.(3)A WiFi and Bluetooth fusion positioning algorithm is designed,which combines the advantages of the two positioning technologies to improve positioning accuracy and stability.Use WiFi and Bluetooth technology respectively for positioning.Each time the WiFi positioning technology outputs a positioning result,the Bluetooth positioning technology outputs multiple positioning results.According to the spatial distance relationship,the Bluetooth positioning results that are beneficial to the positioning accuracy are screened.The adaptive weighting algorithm is used to fuse the positioning results output by the two positioning technologies at the decision layer and output the fusion positioning position coordinates.Compared with Bluetooth positioning technology,the proposed algorithm has increased the probability within 2 m of positioning error from 76.6% to 87.3%.