The complex Swift-Hohenberg equation (CSHE) has been widely studied in recent years. It is a more real model in the mode-locked fiber laser with the saturable absorber due to the addition of the diffraction term compared with the complex Ginzburg-Landau equation. In this work, the dynamics process of the traditional soliton and M−type soliton pulses in the mode-locked fiber laser is demonstrated based on the CSHE model with higher-order nonlinear effects. The results show that with the increase of the small signal gain coefficient, the number of soliton molecules adds gradually. Under the same transmission conditions, the transmission of M−type solitons in the laser are more stable than that of single solitons. By adding the self-steepening effect, it can be found that the time-domain shift due to higher-order dispersion effects is compensated. The self-frequency shift effect caused by the Raman scattering can produce not only time domain shift, but also frequency domain shift. Moreover, the addition of higher-order diffraction term can describe the spectral response of multiple peaks, and makes the pulse spectrum show the asymmetric propagation in the transmission process. Finally, the increase of the length of the single-mode fiber on the right side of the gain fiber in the optical circuit will not only shift the center position of the output pulse backward, but also make the pulse energy show a ladder type downward trend. 

Influence of higher-order nonlinear effects on optical solitons of the complex Swift-Hohenberg model in the mode-locked fiber laser

Multi-layer metal films such as metallic coatings on metal substrate are important elements in modern engineering applications. Specifically, gold-coated metal mirrors are widely used in high-power laser systems. This work studies microscopic energy deposition and transport processes during short-pulse laser heating of multi-layer metals: the absorption of radiation energy by free electrons and the energy exchange between electrons and the lattice. The results show that multi-layer metals present very different thermal responses from single-layer metals during the heating process. In a gold and chromium multi-layer film, although laser energy is absorbed by free electrons in the top gold coating layer, most of the absorbed energy is converted into lattice energy not in the gold layer but rather in the underlying chromium layer. The underlying chromium layer reduces the lattice-temperature rise of the top gold layer significantly during short-pulse laser heating, suggesting a new way to increase the resistance of mirrors to thermal damage in applications of high-power lasers.

Femtosecond laser heating of multi-layer metals--I. analysis

Mode-locked lasers have been widely used to explore interactions between optical solitons, including bound-soliton states that may be regarded as "photonic molecules". Conventional mode-locked lasers normally, however, host at most only a few solitons, which means that stochastic behaviours involving large numbers of solitons cannot easily be studied under controlled experimental conditions. Here we report the use of an optoacoustically mode-locked fibre laser to create hundreds of temporal traps or "reactors" in parallel, within each of which multiple solitons can be isolated and controlled both globally and individually using all-optical methods. We achieve on-demand synthesis and dissociation of soliton molecules within these reactors, in this way unfolding a novel panorama of diverse dynamics in which the statistics of multi-soliton interactions can be studied. The results are of crucial importance in understanding dynamical soliton interactions and may motivate potential applications for all-optical control of ultrafast light fields in optical resonators.

/pdf/synthesis-and-dissociation-of-soliton-molecules-in-parallel-5e2aqyih4s.pdf

Synthesis and dissociation of soliton molecules in parallel optical-soliton reactors

Integrating the information of the first cycle of an optical pulse in a cavity into the input of a neural network, a bidirectional long short-term memory (Bi_LSTM) recurrent neural network (RNN) with an attention mechanism is proposed to predict the dynamics of a soliton from the detuning steady state to the stable mode-locked state. The training and testing are based on two typical nonlinear dynamics: the conventional soliton evolution from various saturation energies and soliton molecule evolution under different group velocity dispersion coefficients of optical fibers. In both cases, the root mean square error (RMSE) for 80% of the test samples is below 15%. In addition, the width of the conventional soliton pulse and the pulse interval of the soliton molecule predicted by the neural network are consistent with the experimental results. These results provide a new insight into the nonlinear dynamics modeling of the ultrafast fiber laser.

Deep neural network for modeling soliton dynamics in the mode-locked laser.

Dissipative systems form various self-organized states owing to the abundant attractor structures. The study of the response of different self-organized states under collision perturbation is of great significance for understanding the dissipative nonlinear systems. The collision dynamics of single soliton and soliton molecules can not only assist the stability analysis of attractors, but also reveal the rich physical connotations of soliton interactions. Here, for the first time, the collision processes of single soliton and soliton molecules in different excited states are detected using the dispersive Fourier transform technology. The collision processes include the disintegration and rebuilding of soliton molecules as well as chaotic oscillating evolution, accompanied by the emergence of transition states such as triple binding state, soliton fusion and acceleration. According to whether the soliton molecule can return to its initial excited state, the collisions are classified as elastic and inelastic. The different interaction strength between solitons is an important condition for rebuilding stable soliton molecules. Numerical simulations show that the gain dynamics are the main physical origin of collisions. Our research will stimulate in-depth research on the interaction of self-organized states in nonlinear systems such as chemical molecules, and have potential applications in optical logic gates.

Elastic and inelastic collision dynamics between soliton molecules and a single soliton.

The dynamics of optical soliton molecules in ultrafast lasers can reveal the intrinsic self-organized characteristics of dissipative systems. The photonic time-stretch dispersive Fourier transformation (TS-DFT) technology provides an effective method to observe the internal motion of soliton molecules real time. However, the evolution of complex soliton molecular structures has not been reconstructed from TS-DFT data satisfactorily. We train a residual convolutional neural network (RCNN) with simulated TS-DFT data and validate it using arbitrarily generated TS-DFT data to retrieve the separation and relative phase of solitons in three- and six-soliton molecules. Then, we use RCNNs to analyze the experimental TS-DFT data of three-soliton molecules in a passive mode-locked laser. The solitons can exhibit different phase evolution processes and have compound vibration frequencies simultaneously. The phase evolutions exhibit behavior consistent with single-shot autocorrelation results. Compared with autocorrelation methods, the RCNN can obtain the actual phase difference and analyze soliton molecules comprising more solitons and almost equally spaced soliton pairs. This study provides an effective method for exploring complex soliton molecule dynamics.

Analysis of real-time spectral interference using a deep neural network to reconstruct multi-soliton dynamics in mode-locked lasers

In passively mode-locked fiber lasers (PMLFLs), the dissipative solitons (DSs) can self-organize to form complex structures through delicate interactions. However, it is still elusive to control these soliton structures by external influences. We here find that at a certain critical power, the location between two soliton molecules can be controlled by a slow modulated pump power. After applying the pump power with periodic fluctuation, two soliton molecules oscillate from the state of soliton molecular complex to stable distribution with maximum inter-molecular separation. During this process, the internal structure of each soliton molecule keeps steady. The slow gain depletion and recovery mechanism which plays a dominant role affects the motion of soliton molecules. These results could further expand the molecular analogy of spectroscopy and stimulate the development of optical information storage and processing.

Oscillatory self-organization dynamics between soliton molecules induced by gain fluctuation.

As a high-throughput data analysis technique, photon time stretching (PTS) is widely used in the monitoring of rare events such as cancer cells, rough waves, and the study of electronic and optical transient dynamics. The PTS technology relies on high-speed data collection, and the large amount of data generated poses a challenge to data storage and real-time processing. Therefore, how to use compatible optical methods to filter and process data in advance is particularly important. The time-lens proposed, based on the duality of time and space as an important data processing method derived from PTS, achieves imaging of time signals by controlling the phase information of the timing signals. In this paper, an optical neural network based on the time-lens (TL-ONN) is proposed, which applies the time-lens to the layer algorithm of the neural network to realize the forward transmission of one-dimensional data. The recognition function of this optical neural network for speech information is verified by simulation, and the test recognition accuracy reaches 95.35%. This architecture can be applied to feature extraction and classification, and is expected to be a breakthrough in detecting rare events such as cancer cell identification and screening.

https://www.mdpi.com/2304-6732/8/3/78/pdf

Optical Machine Learning Using Time-Lens Deep Neural NetWorks

A residual convolutional neural network (RCNN) is introduced to retrieve the separation and relative phase of solitons in soliton molecules in passively mode-locked lasers (PMLs). It proved to be an effective method to explore complex soliton molecule dynamics. © 2020 The Author(s)

Deep neural networks have enabled the reconstruction of optical soliton molecules with more complex structures using the real-time spectral interferences obtained by photonic time-stretch dispersive Fourier transformation (TS-DFT) technology. In this paper, we propose to use three kinds of deep convolution networks (DCNs), including VGG, ResNets, and DenseNets, for revealing internal dynamics evolution of soliton molecules based on the real-time spectral interferences. When analyzing soliton molecules with equidistant composite structures, all three models are effective. The DenseNets with layers of 48 perform the best for extracting the dynamic information of complex five-soliton molecules from TS-DFT data. The mean Pearson correlation coefficient (MPCC) between the predicted results and the real results is about 0.9975. Further, the ResNets in which the MPCC achieves 0.9906 also has the better ability of phase extraction than VGG which the MPCC is about 0.9739. The general applicability is demonstrated for extracting internal information from complex soliton molecule structures with high accuracy. The presented DCNs-based techniques can be employed to explore undiscovered mechanisms underlying the distribution and evolution of large numbers of solitons in dissipative systems in experimental research.

Luhe Zhang

Papers

Analysis of real-time spectral interference using a deep neural network to reconstruct multi-soliton dynamics in mode-locked lasers

Oscillatory self-organization dynamics between soliton molecules induced by gain fluctuation.

Optical Machine Learning Using Time-Lens Deep Neural NetWorks

Analysis of real-time spectral interference using a deep neural network to reconstruct multi-soliton dynamics in mode-locked lasers

Comparing Performance of Deep Convolution Networks in Reconstructing Soliton Molecules Dynamics from Real-Time Spectral Interference