Adaptive mode control of a few-mode fiber by real-time mode decomposition

Mode-modulation-induced high power dual-wavelength generation in a random distributed feedback Raman fiber laser

An all-fiberized random distributed feedback Raman fiber laser (RRFL) with mode-modulation-induced wavelength manipulation and dual-wavelength generation has been demonstrated, where an electrically controlled intra-cavity acoustically-induced fiber grating (AIFG) is employed to adjust the input modal content at the signal wavelength. The wavelength agility of both the Raman effect and the Rayleigh backscattering in RRFL benefits on broadband laser output in case of broadband pumping. The feedback modal content at different wavelengths can be adjusted by AIFG, and then the output spectral manipulation can be ultimately manifested through the mode competition in RRFL. Under the efficient mode modulation, the output spectrum can be continuously tuned from 1124.3 nm to 1133.8 nm with single wavelength, while ulteriorly the dual-wavelength spectrum can be formed at 1124.1 nm and 1134.7 nm with a signal-noise-ratio of 45 dB. Throughout, the power is beyond 47 W with good stability and repeatability. To the best of our knowledge, this is the first dual-wavelength fiber laser based on mode modulation and the highest output power ever reported for an all-fiberized continuous wave dual-wavelength fiber laser.

M2 factor estimation in few-mode fibers based on a shallow neural network

A high-accuracy, high-speed, and low-cost M² factor estimation method for few-mode fibers based on a shallow neural network is presented in this work. Benefiting from the dimensionality reduction technique, which transforms the two-dimension near-field image into a one-dimension vector, a neural network with only two hidden layers can estimate the M² factor directly. In the simulation, the mean estimation error is smaller than 3% even when the mode number increases to 10. The estimation time of 10000 simulation test samples is around 0.16s, which indicates a high potential for real-time applications. The experiment results of 50 samples from the 3-mode fiber have a mean estimation error of 0.86%. The strategies involved in this method can be easily extended to other applications related to laser characterization.

Programmable high-order mode control method based on acoustically induced fiber grating

We demonstrate a programmable high-order mode control method that can be implemented in high-power fiber lasers. 2 W average-power mode-locked pulses are obtained based on a mode-locked fiber laser working in dissipative soliton resonance regime. The fundamental mode (LP₀₁) is fully or partially converted to the high-order modes (LP_11a/b) via an acoustically-induced fiber grating. The mode-superposition fields are recorded using an optical 4f system, and mode components are subsequently analyzed by a mode decomposition algorithm. Our experiments suggest that the mode patterns are stable and dynamically switchable. The method is expected to possess good application value in optical tweezers, fiber communication, laser material processing and other research fields.

Modal decomposition for the fiber beams with arbitrary degree of coherence based on the Wigner distribution function

Modal decomposition (MD) plays an increasingly important role in characterizing fiber beams. Several promising MD techniques have been proposed in literature, all of which are based on a common hypothesis that the modal field is coherently superposed by transverse modes. However, the partially coherent conditions have to be expected in general. In order to take account of this ordinary case, a novel MD scheme employing the Wigner distribution function (WDF) is introduced, which allows the decomposition of fiber beams without any restrictions regarding their degree of coherence. The four-dimensional (4D) WDF (two spatial and two spatial frequency dimensions) of the 2D fiber beam is reconstructed using the coded aperture technique. Based on the measured WDF and orthogonal property of transverse modes, the modal coefficients as well as the mutual modal degree of coherence will be determined unambiguously. The validity and reliability of the proposed approach are illustrated with numerical examples.

Modal decomposition for few mode fibers using the fractional Fourier system

Modal decomposition (MD) has become an indispensable diagnostic tool for optical fibers. A novel MD method using the fractional Fourier system is developed in this paper. Firstly, the existing phase retrieval (PR) algorithm based on the the fractional Fourier transform (FrFT) power spectra is extended to account for the effect of optical vortex. The extended capability can then be employed to reconstruct the phase of modal field to high fidelity in a non-iterative and non-interferometric manner. Combining the reconstructed phase with the measured near-field (NF) intensity, the modal field could be obtained and based upon which the complete MD (involving modal weights and phases) is performed. The validity and reliability of the method are demonstrated through several numerical examples including the noisy signals with different signal-to-noise ratio (SNR) levels.

Learning to decompose the modes in few-mode fibers with deep convolutional neural network

We introduce a deep-learning technique to perform complete mode decomposition for few-mode optical fibers for the first time. Our goal is to learn a fast and accurate mapping from near-field beam patterns to the complete mode coefficients, including both modal amplitudes and phases. We train the convolutional neural network with simulated beam patterns and evaluate the network on both the simulated beam data and the real beam data. In simulated beam data testing, the correlation between the reconstructed and the ideal beam patterns can achieve 0.9993 and 0.995 for 3-mode case and 5-mode case, respectively. While in the real 3-mode beam data testing, the average correlation is 0.9912 and the mode decomposition can be potentially performed at 33 Hz frequency on a graphic processing unit, indicating real-time processing ability. The quantitative evaluations demonstrate the superiority of our deep learning-based approach.

Coupling ultrafast laser pulses into few-mode optical fibers: a numerical study of the spatiotemporal field coupling efficiency

Few-mode optical fibers have been widely used in applications such as data transmission and laser amplification. The coupling of laser beams into such fibers is important because it determines the fiber mode contents, which influence the beam quality, laser amplification efficiency, and appropriate distortion control methods. Ultrafast lasers, as a widely used type of laser source for intense-field applications, can generate pulses with broad spectra and varying phases, adding complexity to the fiber mode coupling problem. In this paper, we numerically study the coupling performance of ultrafast laser pulses into few-mode fibers, investigating the mode coupling efficiency under different launching field conditions. Certain modes can be optimized with their coupling efficiencies in the presence of pulse spatiotemporal couplings, which provides a new perspective toward generating complex waveforms and studying laser matter interactions.

500 W level MOPA laser with switchable output modes based on active control

We report a high power transverse-mode-switchable fiber laser in a master oscillator power amplifier (MOPA) configuration. The output modes of a few-mode fiber amplifier can be actively controlled by the input polarization state of the fundamental mode seed laser using SPGD algorithm. A fast, stable and safety mode switching between LP01 and LP11 modes is achieved in the amplifier at output power of 500 W level.

Multimode fiber modal decomposition based on hybrid genetic global optimization algorithm

Numerical modal decomposition (MD) is an effective approach to reveal modal characteristics in high power fiber lasers. The main challenge is to find a suitable multi-dimensional optimization algorithm to reveal exact superposition of eigenmodes, especially for multimode fiber. A novel hybrid genetic global optimization algorithm, named GA-SPGD, which combines the advantages of genetic algorithm (GA) and stochastic parallel gradient descent (SPGD) algorithm, is firstly proposed to reduce local minima possibilities caused by sensitivity to initial values. Firstly, GA is applied to search the rough global optimization position based on near- and far-field intensity distribution with high accuracy. Upon those initial values, SPGD algorithm is afterwards used to find the exact optimization values based on near-field intensity distribution with fast convergence speed. Numerical simulations validate the feasibility and reliability.