Research And Implementation Of Identity Recognition Method Based On WiFi Channel Characteristics

The issue of human identity recognition has been studied for many years.With the advent of the information age,people are increasingly facing identification problems in their daily lives.Human identity in the information age is a technology which collects the physiological or behavioral characteristics inherent in the human body by computer and thereby performs personal identification.There are many methods for human recognition in the industry,such as face recognition,iris recognition and fingerprint recognition.These methods have corresponding advantages and disadvantages in their respective application scenarios.At the same time,the development of wireless communication technology has brought new enlightenment to non-invasive biometric technology.Specifically,channel state information(CSI)can effectively infer movement according to signal influences generated by human actions,and modern WiFi standards such as IEEE 802.11n/ac transmit a large amount of channel state information data during communication,the received multi-channel signal is the result of gain and attenuation after multiple reflections by surrounding obstacles.This thesis has conducted in-depth research on the above mentioned issues.Specifically,the current human identification technology is sorted out,and the latest related work and models are summarized.The technical principles,application scenarios,advantages and disadvantages of different human identification technologies are analyzed and discussed.Furthermore,this thesis investigates the principle and application of channel state information of fine-grained information of WIFI,mainly focusing on how the channel state information reflects the combined effect of surrounding objects.On this basis,this thesis mainly designs and verifies the three aspects of the WiFi-based human identity identification method:(1)The principle of channel state information and its application technology research,(2)Research on human identification technology and research on identity technology using WiFi(3)Feasibility Study of Human Identity Recognition System Based on WiFi.Based on the above related research,this thesis proposes a novel human identification method.Specifically,it is observed that each person has a unique influence on the channel state information of the surrounding WiFi when moving indoors,and extracts these effects as a database of respective feature data by using various signal processing methods,and finally combines with the machine learning method.Training and forecasting to achieve the purpose of identification.Finally,simulation and actual test experiments were carried out for the design of this thesis,and the experimental results were analyzed.According to the comparison results,the design of this work can identify the human identity in the environment with WiFi.Under the condition of 6 participants,the correct recognition rate of identity can reach more than 90%.In small environment(Family,office and so on),human identification method in this thesis has the advantages of low cost,low environmental impact,no need to introduce additional equipment and non-infringement of user privacy,compared with traditional identification methods based on biometrics or wearable devices.