Research On Suspicious Liquid Detection In Multi-band Based On Channel State Information

In the process of security inspection of various projects in public places,the detection of dangerous liquid is an important work.At present,the most commonly used technology in the field of safety inspection in China is Spectroscopic Computed Tomography(SPC).By detecting the dielectric constant and conductivity of the liquid,according to the large difference between the dielectric constant and conductivity of the safe liquid and the dangerous liquid,we can judge whether the liquid is flammable or explosive.Although the existing liquid detection technology is very mature,it is not conducive to large-scale deployment because of its high price and large volume.In view of the above problems,and combined with the rapid development of Wi-Fi technology,this thesis proposes a method based on Channel State Information(CSI)technology to detect the dangerous liquid.This method uses the difference of dielectric constant between different liquids,and takes the difference of CSI amplitude data before and after the Wi-Fi signal passes through the liquid as the judgment basis.This method has the advantages of low price,easy deployment,good accuracy and robustness.In addition,on the basis of CSI amplitude data,the results of liquid detection using C-band wireless signal and S-band wireless signal are discussed,as well as the effects of different containers on liquid detection results.In the data extraction part of CSI,firstly,the thesis introduces the experimental platform and the design of the whole experimental process;secondly,a CSI data extraction model is proposed;finally,the feasibility analysis of liquid detection using wireless signal and CSI technology is realized,and a variety of experiments of liquid detection and recognition are completed,that is,the experiment of judging suspicious liquid and nonsuspicious liquid;the experiment of identifying suspicious liquid;the experiment of identifying different concentrations of alcohol.The CSI amplitude data of different liquids are collected,which provides theoretical basis and data source for liquid classification.In the part of liquid identification,firstly,Hampel filter is used to eliminate the environmental noise and data singularity;secondly,the Principal Component Analysis(PCA)algorithm is used to reduce the dimension and select the feature of CSI data set;finally,K-Nearest Neighbor(KNN)classification algorithm and Random Forest(RF)classification algorithm are compared for liquid classification results,to select the optimal model and optimal parameters.The experimental results show that the method proposed in this thesis can detect the suspicious liquid with the accuracy of more than 98%,the recognition accuracy of the suspicious liquid type is more than 97%,and the recognition accuracy of different concentrations of alcohol is as high as 98%.In addition,on the basis of CSI amplitude data,the results of liquid detection using C-band wireless signal and S-band wireless signal are discussed,as well as the effects of different containers on liquid detection results.