Research On Indoor Multi-Scenario Fingerprint Localization Algorithm Based On CSI

With the popularization development of intelligent devices combined with various computer technologies and the extensive deployment of mobile Internet,location-based services have also been developed rapidly.At the same time,location service has become an indispensable part of people's lives,so many researchers have proposed different methods based on Wi-Fi,Bluetooth,ultra-wideband technology and RFID technology try to provide location services in an indoor environment.Among these methods,the method based on Wi-Fi equipment has the characteristics of low cost,easy deployment and high scalability,which makes the indoor positioning system based on Wi-Fi become the research hotspot in recent years.In the traditional Wi-Fi indoor location method,the received signal strength indication is often used as fingerprint feature,but RSSI is very unstable,it is very vulnerable to environmental noise interference,so it will produce biased positioning results,at the same time,It also affects the overall accuracy.Especially in the multi-scene indoor environment,the application of indoor location method can't be extended.With the support of orthogonal frequency division multiplexing(OFDM)and multiple-input multiple-output(MIMO)technology,we can obtain CSI signals from Wi-Fi devices and open up a new indoor location technology by using finer-grained CSI features.In this paper,we will use CSI to study the location of indoor multi-scene.The main research content and work of this paper are summarized as follows:(1)First of all,this paper introduces the background and technical characteristics of different indoor positioning technology,and compares the advantages and disadvantages of CSI and RSSI signals as fingerprint information.In addition,in order to achieve high-precision location algorithm,we propose a passive indoor fingerprint algorithm based on CSI.The algorithm establishes the relationship between the change characteristics of CSI signal and the location of the target.through the CSI data collected by Wi-Fi equipment,the principal component analysis method is used to process the CSI amplitude data,and the main data features are extracted as fingerprint database.Finally,the naive Bayesian classification algorithm is used for real-time location.This method further improves the efficiency and greatly reduces the influence of multipath effect on the positioning results when RSSI signals are used.(2)In view of the complexity of indoor environment,variability,occlusion of signal transmission and many other challenges faced by CSI-based indoor positioning,we propose a location algorithm that combines the amplitude and phase information of CSI.In this algorithm,the CSI phase feature signal is processed by linear transformation and fused with the filtered amplitude information to build a more robust fingerprint feature.It not only weakens the influence of unstable factors in indoor environment on the positioning results,but also makes full use of the rich signal characteristics provided by CSI to open the application scene from the experiment to the specific application scene.(3)Considering the complexity and uncertainty of some indoor special·environments,we analyzed the error of CSI phase data,and proposed a positioning algorithm in complex indoor environment.This algorithm solves the influence of multipath effect and noise interference on the positioning accuracy of CSI phase signal in complex environment.At the same time,it also improves the positioning accuracy of CSI positioning algorithm in complex indoor environment.So that the application scenario of the algorithm is more extensive.