Search

In this paper, a simple implementation of dual-channel chaotic lasing is realized using a fiber ring laser with an intra-cavity semiconductor optical amplifier (SOA). Due to the high-speed gain saturation and recovery, combined with the strong coupling between the amplitude and the phase in the SOA, chaotic oscillation can be facilitated. Meanwhile, the Kerr effect and the anomalous dispersion in the single-mode fiber causes modulational instability, which strengthens the chaotic behavior, leading to reduced chaotic oscillation threshold. The output optical signals with orthogonal polarizations are split using a polarization beam splitter and detected in parallel by two photodetectors (PD). The electrical outputs are acquired by a real-time oscilloscope with an 8-bit ADC sampling at a rate of 80 GSa/s. Both polarizations exhibit electrical bandwidth up to 13 GHz, which is limited by the analog bandwidth of the PD. While the autocorrelation of each signal shows a typical curve of a random signal, the cross-correlation of the two polarizations contains no significant correlation peak, indicating that the two chaotic signals are independent of each other. Random number generation at a bit rate of 480 Gb/s is then realized using the acquired sequences. The random bits can pass all the randomness tests provided by the NIST SP800-22 test tool. Due to the independence of the two polarizations, this chaotic laser also serves as a promising candidate for chaotic communication exploring polarization division multiplexing.

By the Monte Carlo method it has been shown that the molecular conductance of an elbow, the correction factor and the transmission probability are all functions of the supplementary angle between the two tubes. These three quantities at first decrease slowly with increasing angle, pass through a minimum between 49° and 73° and then again increase. Our calculations differ from the results of Roth who predicted a monotonous decrease in the transmission probabilities with angle and of Davis who stated that the transmission probabilities at 90° and 0° were equal. Using a non-linear regression technique, an empirical expression for the correction factor K of the elbow as a function of the angle θ between the two tubes has been fitted to the calculated results. The expression for K has the form K = A + Be Cθ + D sin E (θ + F) where A, B, C, D, E and F are the regression coefficients.