Research And Implementation Of Access Authentication System For Iot Terminal Based On Device Fingerprint

The technology of Internet of things(IoT)has get rapid development in recent years,and with the deployment of massive IoT terminals,the security problems of terminals emerge in endlessly.The access authentication of IoT terminals is very important in the security of the Internet of things.Device fingerprint is one of the common access authentication methods in the IoT environment.Device fingerprint can be divided into active device fingerprint and passive device fingerprint.The device recognition rate of active is higher than passive method.But the dominant and implicit features are extracted from the device itself,which has the problem that some IoT terminals do not support and these features have the risk of forgery.However,the passive fingerprint features,they are all come from the traffic data of the terminal device,all terminals are available.In order to complete the access authentication better,this paper uses the idea of two factor authentication(2FA),and proposes a method to authenticate the device identity by using active and passive device fingerprint.At present,active device fingerprint is relatively mature,but the existing passive device fingerprint still needs to be improved.Therefore,this paper proposes an IoT terminal access authentication algorithm based on deep learning.The algorithm extracts features from traffic data to construct device fingerprint,uses fixed window sliding sampling mechanism to solve the problem of data dimension consistency and interference vector,adopts SMOTE algorithm to solve the problem of sample imbalance,and finally uses bidirectional long-term memory network to realize device identification.Simulation results show that,compared with traditional machine learning and LSTM algorithm,the recognition accuracy of algorithm which proposed by this paper is improved by at least 6%.Based on the above,this paper designs and implements an IoT terminal access authentication system based on device fingerprint.The system can use active and passive fingerprint to complete the access authentication of IoT terminal.After completing the functional requirements analysis and functional architecture design of the system,this paper designs and implements each module of system.Finally,this paper tests the access authentication system.The test result shows it is consistent with the design description,and its average transaction response time of the system is still less than ls when hundreds number of IoT terminal sends requests to the system at the same time.