Search

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 evaluate a method that could be tried by an eavesdropper to force the standard two-laser scheme of chaos private transmission. This method is based on the detection of the voltage arising across a semiconductor laser junction and uses two lasers: one is synchronized to the transmitter (thus, reproducing chaos and rejecting the message), the other is selected and trimmed to copy the transmitter (i.e., reproducing both chaos and the message). The message is then obtained by difference and chaos cancellation. By assuming the usual parameter mismatch, which is expected to be achieved by an eavesdropper, it is shown that the method does not represent a substantial threat to the privacy of the transmission system. The scheme is also briefly considered as an alternative, but less performant, scheme for chaotic cryptography. The simulations are based on the Lang–Kobayashi model. [ABSTRACT FROM AUTHOR]

In this paper, we compare the performances of the well-known two-laser and three-laser chaos communications schemes, in terms of privacy. It is found by numerical analysis that in optimized conditions both methods are well suited for private transmission. However, while the three-laser scheme gives its optimum performance with the twin laser pair, the two-laser scheme reaches its best performance using two slightly different lasers (the optimum pair). This results in an easier device selection and in a slightly higher privacy level of the three-laser scheme. Using a twin pair with the two-laser scheme still offers a good level of privacy in most cases; however, in principle the eavesdropper has a chance to extract the message, by selecting a suitable laser. [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]