Research And Application Of Indoor Positioning Algorithm Based On WLAN Location Fingerprint

With the widespread popularity of wireless networks and the rapid development of mobile terminals,the demand of mobile users for location information extends from outdoor to indoor.However,GPS can't achieve effective positioning in the interior and urgently needs effective indoor positioning technology to cope with people's growing indoor positioning needs.WLAN positioning system based on received signal strength,full use of the existing wireless network infrastructure,without additional dedicated hardware device,can be implemented in pure software positioned in any way an intelligent terminal having a Wi-Fi adapter.In this thesis,we analyze the RSS features and optimize some existing problems of location fingerprinting algorithms.This thesis proposes a WLAN indoor positioning algorithm based on space partitioning technology for PCA fingerprinting and designs and develops an indoor positioning application system.The main work of this thesis is as follows:(1)Aiming at the instability of WLAN location fingerprinting caused by the difference of received signal capabilities of heterogeneous terminals,the spatial characteristics and variation of RSS are studied.Designs RSS error correction algorithm for heterogeneous terminals,correct the RSS value while positioning online to reduce the positioning error caused by the inconsistency between the terminal building fingerprint library and real-time positioning.(2)In order to solve the problem that the location fingerprint database contains a lot of redundant and noisy information and the location fingerprint information is over-dimensioned,the principal component analysis(PCA)method is used to extract and reduce the dimensionality of the location fingerprint so as to reduce the impact of fingerprint redundancy and multipath effects RSS value fluctuates.The PCA fingerprinting method is designed to increase the PCA fingerprinting coefficient and reduce the positioning error caused by the traditional fingerprinting algorithm which only uses the RSS Euclidean distance between the target point and the reference point as the similarity criterion.Experiments show that the proposed algorithm can effectively extract and reduce the dimensionality of location fingerprints,suppress the influence of ambient noise to a certain extent,improve the positioning accuracy and save the fingerprint storage space.(3)In order to solve the problem of too high computational complexity of fingerprint database matching in online positioning,a location partition algorithm of location fingerprinting library is designed to cluster and segment the location fingerprint database and reduce the fingerprint searching space in the matching computing process.The Gap statistic method is used to improve the K-value of K-means algorithm,which is used to determine the optimal initial clustering number K to avoid the positioning error caused by the uncertainty of K value.The optimization initial clustering center selection strategy is used to select Uniformly distributed initial cluster centers to avoid falling into the local optimum and result in invalid clustering,thereby reducing the complexity of the fingerprint matching calculation in online positioning.