Effect of feed-forward in chaotic synchronization matching between master laser diode and two slaves

A minor laser network in this experimental study is constructed to simulate a chaotic neural network. The system includes a master laser, which is feeding two slaves via the feed-forward technique. This technique includes variations in modulation strength for the master laser signal. While the two slaves were subjected to both modulation from the master and self-optoelectronic feedback. The results confirm the existence of identical synchronization (S) between each of the two laser pairs in the network. Correlation coefficient between master laser and slaves fluctuated from 0.70 in slave laser two (SL2) to 0.99 for slave laser one (SL2). Linearity between those laser pairs is calculated statistically in addition to measurements for Fourier spaces and attractors to ensure identical pulsation for lasers, achieving values of 1.30 SL1 and 1.12 SL2 with the master laser, respectively. Results indicated that the three LDs in the system are all emitting at identical amplitudes. Such observation confirms optimal identical synchronization for laser current, feedback modulation RF, and optical strengths as initial conditions. Measured slopes for resulted amplitudes between master and slave lasers in the system are affected by lag and optimized to increase their values. Sin function alteration is observed for the number of spikes when chaotic emission is established. This indicated that the system is altered from one stability state to a different one and then to a chaotic one based on selected initial conditions in a periodic manner. Similarity and matching between these laser emissions are the main requirements for encrypted chaotic communications, in particular for optical neural networks.