Research On Identity Authentication In Physical Layer Of Passive Optical Network Based On Hardware Fingerprint

Information society has put forward higher and higher requirements for information security.As the key technology of "last kilometer" connecting users and backbone networks,optical access network technology has become a hot topic and focus of research.As a kind of physical layer device information,hardware fingerprint is unique,difficult to forge and widely exists in all kinds of hardware devices.Therefore,identity authentication technology based on hardware fingerprint has high security.At the same time,compared with other physical layer authentication technology,hardware fingerprint technology does not need to design complex security algorithm or specific encryption strategy,which makes the identity security technology based on hardware fingerprint easier to implement.However,the existing research on physical layer hardware fingerprint of optical access network is inadequate.In this paper,the hardware fingerprint extraction and identification method are designed and verified.The main research contents are as follows:(1)Aiming at the problem that existing algorithms need to extract features based on the pilots of orthogonal frequency division multiplexing signal and have high complexity,a hardware fingerprint feature extraction method based on channel noise model is proposed,and the algorithm is applied to extract the fingerprint of optical network in the OFDM passive optical network.Then,the convolutional neural network is used to classify the extracted features,and the hardware fingerprint identification without pilot assistance is successfully realized.Compared with the existing methods,the recognition accuracy is higher.The recognition accuracy of the legal terminal is 99.25%and the recognition accuracy of identification of illegal terminal is 100%?(2)The existing research only realized hardware fingerprint extraction and recognition based on orthogonal frequency division multiplexing signal.In this paper,the fingerprint features carried by binary amplitude shift keying signal are extracted based on wavelet decomposition and reconstruction feature extraction algorithm,and its effectiveness is verified.In this paper,experiments are carried out from five aspects:sample data size,wavelet decomposition and reconstruction method,wavelet generating function,number of terminals to be identified and optical fiber channel transmission,and the law of recognition rate affected by the above factors is studied.