Design And Implementation Of Fine-Grained Indoor Ositioning System Based On Informer

In recent years,with the increasing popularization of wireless networks,the application of wireless fidelity (Wi-Fi) positioning technology has become more and more extensive,especially the emergence of fine-grained Channel State Information (CSI) signals has given it greater a competitive advantage in the positioning field.Currently,most indoor positioning platforms still focus on single-floor positioning,however,facing the rising multi-story buildings,it is obviously already difficult to satisfy the positioning requirements of users with only twodimensional plane location information.To address the above positioning requirements,this paper conducts an investigation into current indoor positioning technologies,summarizes the drawbacks of existing methods,thus clarifies the implementation of the algorithm.This paper utilizes Received Signal Strength Indicator (RSSI)and CSI signals extracted from Wi-Fi signals to implement the positioning algorithm and develops an indoor positioning system suitable for multifloor application scenarios,which effectively compensating for the shortcomings of current indoor positioning platforms.This paper implements floor recognition algorithm based on RSSI and intra-floor positioning algorithm based on CSI using the Informer model.Through combining the above two algorithms,this paper achieves positioning in multi-floor buildings.The algorithm proposed in this paper can not only handle long sequences,but also optimize the computational cost caused by too much fingerprint data.After completing the algorithm,an indoor positioning system has been developed.In terms of the process of software development,this paper first carries out a user-oriented demand analysis to clarify the system’s required functionalities.Then,it completes the outline design and establishes the logical model,and finally explains the implement procedure of each functionality in turn exhaustively through the detailed design.After developing of the system,this paper conducts a series of tests to ensure that the system can achieve the expected functionalities and work normally.This paper collects RSSI data from three different scenarios:Wucai City,Xidan Joy City and Zhuoyuehui.This paper tests the proposed floor recognition algorithm,and the overall accuracy is over 95%.This paper tests the intra-floor positioning algorithm on the IPIN2021 track7 dataset,and the average positioning error is only 0.732 meters.The accuracy of the above two algorithms can meet the requirements of users,and the effectiveness of the proposed algorithm has been verified through concrete experimental results.This paper provides an effective solution for positioning in multi-story buildings,and effectively improves the positioning accuracy using fine-grained CSI signals.