Search

We numerically study a new hardware method for network authentication, using a pair of chaotic lasers subject to the same optical injection from a third chaotic laser, in a close-loop configuration. One laser is in the secure environment, the other is in the unsecure environment, and if their parameters are matched (i.e., they are 'twins'), in suitable operating conditions they synchronize and generate the same pseudo-random amplitude modulation. Access to the secure environment is authorized only if the two responses are found equal. As in other schemes based on Physical Unclonable Functions (PUFs), security relies on unavoidable small differences between nominally identical but physically separate devices: due to the strong sensitivity of chaos to laser parameter dispersion, finding a suitable laser to synchronize is a very difficult task for the Adversary, while the laser of the Authorized user can be selected as a twin of the one in the secure environment. [ABSTRACT FROM AUTHOR]

In this paper, we numerically evaluate a hardware-based method for authentication using a pair of twin chaotic lasers. This method belongs to the PUF (Physically Unclonable Function) class; however, it does not require storing the response of the hardware in a database, because both users own the same hardware (one of the twin chaotic lasers), producing, on the fly, the same response, i.e., the same chaotic waveform, when subject to the same stimulus (injection from a third laser) in the same working conditions (local injection, pump current). A bit sequence can be easily obtained from the chaotic waveforms by electronic processing, and authentication consists in comparing the sequences produced by the twin lasers. This scheme is proposed above all for authentication of a client in an unsecure environment to a server in a secure environment, but it can be used also for peer-to-peer authentication. Both the case of open- and close-loop chaotic lasers are considered. Simulations are based on the Lang-Kobayashi model. [ABSTRACT FROM AUTHOR]

In this paper, we numerically demonstrate that the performances of private transmission using chaotic lasers can be substantially improved by double masking the message by a suitable scheme. In addition to standard optical chaos, our method makes use of the chaotic voltage waveform available across the transmitter laser junction. With respect to the standard setup, the new one requires only one more electrical amplifier and a passive adder at the transmitter, while the receiver remains unchanged. It has been found that with this new scheme, synchronization quality is significantly improved for the twin lasers, while this is not the case for an unmatched laser pair. This results in an increased bit error rate differential between the authorized user and an eavesdropper, thus enhancing privacy. Simulations have been performed with the Lang–Kobayashi model. [ABSTRACT FROM AUTHOR]