Lax-integrable partial differential equations (PDEs) can by definition be described through a compatibility condition between two linear operators. These operators are said to form a Lax pair for the PDE, which itself is usually nonlinear. Lax pairs are a very useful tool, but unfortunately finding them is a difficult problem in practice. In this paper, we propose a method that determines the spectral operator of an AKNS-type Lax pair such that the corresponding PDE fits given measurement data as well as possible. The spectral operator then enables practitioners to solve or analyze the underlying PDE using the induced nonlinear Fourier transform. The underlying PDE only has to be approximately Lax-integrable; the method will find the spectral operator that explains the data best. Together with the dispersion relation, the spectral operator of AKNS type completely determines an integrable PDE that approximates the true underlying PDE. We identify the most suitable spectral operator by matching PDE-dependent quantities that should be conserved during evolution. The method is automatic and only requires recordings of solutions at two different values of the evolution variable, which do not have to be close.

We experimentally investigate the problem of monitoring the Kerr-nonlinearity coefficient $\gamma$ from transmitted and received data for a single-mode fiber link of 1600 km length. We compare the accuracy and speed of three different approaches. First, a standard split-step Fourier method is used to predict the output at various $\gamma$ values, which are then compared to the measured output. Second, a recently proposed nonlinear Fourier transform (NFT)-based method, which matches solitonic eigenvalues in the transmitted and received signals for various $\gamma$ values. Third, a novel fast version of the NFT-based method, which only matches the highest few eigenvalues. Although the NFT-based methods do not scale with link length, we demonstrate that the SSFM-based method is significantly faster than the basic NFT-based method for the considered link of 1600 km, and outperforms even the faster version. However, for a simulated link of 8000 km, the fast NFT-based method is shown to be faster than the SSMF-based method, although at the cost of a small loss in accuracy.

We present a fast method to calculate the significantly large solitonic components of signals with large time-bandwidth products governed by the nonlinear Schrödinger equation, for which the computation typically becomes prohibitively expensive and/or numerically unstable. We partition the full signal in both frequency and time to obtain short signals with a constant number of samples, independent of the size of the full signal. The solitons within each short signal are computed using a conventional nonlinear Fourier transform (NFT) algorithm. The partitioning in general leads to spurious solitons not present in the full signal. We therefore design an acceptance scheme that removes spurious solitons. The remaining solitons are attributed to the full signal. Solitons that are too wide to fit into the short signals cannot be detected by this approach, but since wide solitons must be of low amplitude, the significant solitons will be found. This approach only requires O(N) floating point operations, with N the number of signal samples. It can furthermore be applied to signals with large time-bandwidth products for which conventional NFT algorithms become unreliable or even fail. When applying our proposed method to a signal of 15,000 samples, the significant solitonic components were computed 14 times faster than when considering the whole signal, for which the conventional algorithm furthermore provided wrong results. We found that time-partitioning yields accurate results, while frequency-partitioning causes a small loss in accuracy. Combined frequency-time partitioning leads to the fastest computation, but also suffers from the same loss in accuracy as with frequency-partitioning. As time-partitioning yields a significant speed-up at nearly no loss in accuracy, we regard this as the method of choice in most practical scenarios.

Recently, a nonlinear Fourier transform-based Kerr-nonlinearity identification algorithm was demonstrated for a 1000 km NZDSF link with accuracy of 75%. Here, we demonstrate an accuracy of 99% over 1600 km SSMF. Reasons for improved accuracy are discussed.

Efficient fiber-optic communication requires precise knowledge of the fiber coefficients, but these often change over time due to factors such as aging or bending. We propose a novel algorithm that identifies the average second-order dispersion and Kerr nonlinearity coefficient of a fiber, without employing any special training signals. Instead, ordinary input and output data recorded during normal operation is used. To the best of our knowledge, this is the first such algorithm. The algorithm is based on the nonlinear Fourier spectrum of the signal, which is known to evolve trivially as the signal propagates through an idealized model of the fiber. The algorithm varies the values of the fiber coefficients until the corresponding nonlinear Fourier spectrum at transmitter and receiver match optimally. We test the algorithm on simulated transmission data over a 1600 km link, and accurately identify the fiber coefficients. The identification algorithm is in particular well suited for providing a fiber model for nonlinear Fourier transform-based communication.

The Material Point Method (MPM) is a numerical technique that combines a fixed Eulerian background grid and Lagrangian point masses to simulate materials which undergo large deformations. Within the original MPM, discontinuous gradients of the piecewise-linear basis functions lead to the so-called grid-crossing errors when particles cross element boundaries. Previous research has shown that B-spline MPM (BSMPM) is a viable alternative not only to MPM, but also to more advanced versions of the method that are designed to reduce the grid-crossing errors. In contrast to many other MPM-related methods, BSMPM has been used exclusively on structured rectangular domains, considerably limiting its range of applicability. In this paper, we present an extension of BSMPM to unstructured triangulations. The proposed approach combines MPM with C¹-continuous high-order Powell–Sabin spline basis functions. Numerical results demonstrate the potential of these basis functions within MPM in terms of grid-crossing-error elimination and higher-order convergence.