Key Technologies Of High-precision Secure Indoor Positioning Based On Physical Layer Parameters In Complex Environments

With the advent of the Internet big data era and 5G mobile communications,locationbased services have become increasingly important.In recent years,some studies have focused on how to accurately obtain the location information of target.In the actual environment,the complexity of wireless positioning depends on the actual situation of the area.According to the current environment and the physical location of the deployed access device and the target to be located,it is divided into line-of-sight,non-line-of-sight and Complex mixed scenes.Nowadays,location information is private in the era of Internet big data,and a certain security and confidentiality mechanism is required.Therefore,it is necessary to develop a high-precision,safe and reliable positioning system solution.First,this article describes some methods and technologies for obtaining physical location information,including infrared,Bluetooth,RFID,UWB,and Wi-Fi,etc.,and analyzes the advantages and disadvantages of each method.Aiming at the current actual deployment and general application level,this article focuses on the research and analysis of the development of Wi-Fi positioning technology and focuses on effective positioning methods for complex environments,including the improvement of fingerprint-based positioning technology and the security problems in the positioning environment.Secondly,aiming at the positioning accuracy problem in the indoor complex environment,this paper focuses on the single AP fingerprint-based positioning system on CSI.The EDBFi positioning system proposed in this paper can solve the problem of low accuracy of traditional fingerprint algorithms and small granularity of position fingerprints in complex environments.After verification by the actual hardware platform,the accuracy of the classic fingerprint positioning system Deep Fi is increased by 20%.The average positioning error is about 65 cm.Specifically,the EDBFi positioning system mainly includes two stages.The offline stage uses the EDBN network to reconstruct the location fingerprint to increase the granularity.Besides,the online stage uses a high-dimensional matching function to convert the correlation between locations into a similarity probability.In this paper,the EDBFi positioning system is verified by the CSI obtained through the actual hardware test by building an experimental platform,and the influence of the number of antennas,matching function and other parameters on the performance of the EDBFi positioning system is analyzed.Secondly,aiming at the problem of positioning accuracy in complex indoor environments,this paper proposes a single AP fingerprint database positioning system based on CSI through research.The fingerprint database positioning system based on EDBN proposed in this paper can solve the problem of low accuracy of traditional fingerprint algorithms and small fingerprint size in complex environments.After verification by the actual positioning platform built,it is improved compared with the classic positioning system Deep Fi.With an accuracy of 20%,the average positioning error is about 65 cm.Specifically,the fingerprint positioning system in this paper is divided into two stages.The offline stage uses the EDBN network to reconstruct the position fingerprint to increase the granularity,and the online stage converts the position similarity probability through a high-dimensional matching function.Finally,an experimental platform was built to verify the CSI of the actual hardware test,and the influence of the number of antennas,matching function and other parameters on the positioning accuracy was analyzed.Finally,this article studies the security and credibility problems in the actual positioning system.First,a basic attack detection model based on RSS is established,and the distribution of RSS in the presence and absence of spoofing attacks is analyzed.Secondly,the article introduces the traditional spoofing attack detection model based on K-means clustering algorithm and analyzes its defects,so this article proposes a spoofing attack detection model based on PKC algorithm.The detection rate and false alarm rate of the detection model proposed in this paper and the traditional detection model are measured by simulating the possible deception attacks of the actual positioning system,and the detection model is analyzed from different parameter dimensions such as parameter learning rate and inertial weight.The simulation results show that the detection model based on the PKC algorithm proposed in this paper has a better detection rate under different decision thresholds.