Signal Fingerprint Identification Of Wireless Communication Equipment Based On Transient Analysis

With the development of science and technology,modern society is gradually becoming an information age.The wireless communication technology,which is widely used in military and people's lives,has provided a variety of communication ways for all of us.This technology,combined with the Internet,has also brought earth-shaking changes to human life.With the continuous development of technologies and the improvement of user needs,the security of wireless communication has become a new challenge.Usually,wireless communication network protects communication by using security protocol based on key mechanism.But with the continuous development of hacking technology,it cannot protect the communication security effectively.However,it will work if we combine non-copyable fingerprint characteristics with encryption key as the way of authentication.In addition,with the continuous integration of communications technologies into the military field,the military battlefield also puts forward higher demands for the reconnaissance and individual identification functions of combat communications equipment.This paper deeply studies the transient signal fingerprint recognition of wireless communication devices,including the detection of the transient starting point,feature extraction and the design of recognizer.The main points are as follows:1.This paper builds a model framework of the transient signal fingerprint recognition system of wireless communication devices,which mainly includes four modules: signal acquisition,transient signal extraction,fingerprint characteristics extraction and fingerprint characteristics matching.2.This paper put forwards a method of transient starting point detection based on fuzzy entropy.Compared with phase detection and short-term energy detection,the transient starting point detection based on fuzzy entropy is more effective in detecting leading response part accurately.3.This paper put forwards a method of the transient signal fingerprint characteristics extraction based on fuzzy entropy.Compared with the least-squares polynomial fitting feature,it has higher discrimination between the different models of the same-brand mobile phones or different brands.The two features adopt integration of the line fusion method,which achieve better recognition results.The average recognition rate of 4 same-brand mobile phones of the same model(70 signal samples per mobile phone,280 signal samples in total)is 96.25%.4.Two feature recognizers(BP neural network and PNN probability neural network)are designed to perform individual identification experiments on transient fingerprint characteristics.The BP neural network performs recognition experiments on the fusion features of the same-brand mobile phones with the same models,the same-brand mobile phones with the different models and different brands of mobile phones.The average recognition rates are 96.25%,98.5% and 99.44% respectively.It can be seen that the fusion of the fuzzy entropy feature and the least-squares polynomial fitting feature can effectively achieve the individual identification of mobile phones.