Research Of Wi-Fi Fingerprints Optimization And Indoor Map Construction Algorithm

Today,more and more applications provide services based on location.The market has huge potential in location-based services and real-time location system.Academia and industry are committed to solving the technical problems of indoor positioning,with various solutions emerging.Wi-Fi positioning system is easier to operate and low-cost,due to its widely distributed infrastructure and mobile devices that are commonly carried by the general public.However,Wi-Fi positioning also faces a series of technical problems.For Wi-Fi fingerprint positioning system,higher positioning precision requires an accurate fingerprint database;however,its acquisition is labor-intensive.Although the Wi-Fi SALM method widely studied at present can reduce the manpower and time cost of the fingerprint collection,it also sacrifices the accuracy of the fingerprint database,and the positioning accuracy decreases accordingly.Therefore,the collection,maintenance and optimization of the fingerprint database in Wi-Fi positioning has always been a research hotspot.Besides,the systems for performing indoor positioning often requires indoor map in advance,but the mapping is a significant amount of work.The existing map construction methods including vision,laser radar,sonar,etc.,are either more influenced by the environment or have higher costs,with little implementation.For the noise problem of fingerprint database,a parameterized fingerprint construction system is proposed to obtain an accurate fingerprint database.In the proposed scheme,Wi-Fi signal strength is collected in an INS-aided fingerprint collection system along pre-defined path to associate each fingerprint with its location automatically instead of the attachment of labels manually which makes the collection process more efficient.The wall-induced wireless signal propagation theory taking the floor plans into account is used to model the AP in the area of interest by using the least squares fitting method.Finally,a fingerprint database is constructed with the assistance of RSS measurements that matches the surrounding environment.An experiment was carried out at the Haiyun teaching building of Xiamen University.Compared with the original fingerprints,the positioning accuracy of the fingerprints obtained by the proposed scheme is improved,on the average,by 0.6 meters according to the positioning results from many tests.Aimed at the problem of complex progress for indoor map building,the paper proposes an RSS-based indoor map construction algorithm using the fingerprint and its location information in the original fingerprint database.The indoor map construction problem is transformed into the classification problem of reference points in fingerprint database.With the aid of hierarchical classification system consisting of single-AP-based base classifier and multi-AP-based combination classifier,the accuracy of the classification results is guaranteed.And an accurate line segment feature map is obtained by identifying demarcation line between two classes from the hierarchical classification system.At the same time,the theory of wireless signal propagation is used to model AP and find out the abnormal environmental factors to detect the presence of walls.The simulation experiments are used to verify the effectiveness of the classification algorithm and the feasibility of the map construction algorithm.The results of field experiments are also presented in the last.The fingerprint database optimization algorithm can also be extended to BLE and other wireless signal-based fingerprint positioning system.And map construction method is a very practical technology,with extensive application.