A5075626181- Optical Network Technologies
- Advanced Photonic Communication Systems
- Advanced Fiber Laser Technologies
- Photonic and Optical Devices
- Neural Networks and Reservoir Computing
- Photonic Crystal and Fiber Optics
- Semiconductor Lasers and Optical Devices
- Advanced Fiber Optic Sensors
- Nonlinear Photonic Systems
- Neural Networks and Applications
- Advanced Optical Network Technologies
- Neural dynamics and brain function
- Cognitive Radio Networks and Spectrum Sensing
- Optical and Acousto-Optic Technologies
- Distributed Sensor Networks and Detection Algorithms
- Laser Design and Applications
- Wireless Communication Networks Research
- stochastic dynamics and bifurcation
- Sparse and Compressive Sensing Techniques
- Optical Imaging and Spectroscopy Techniques
- Cybersecurity and Information Systems
- Advanced Optical Sensing Technologies
- Optical Systems and Laser Technology
- Advanced Control and Stabilization in Aerospace Systems
- Optical Wireless Communication Technologies
Aston University
2015-2024
Yazd University
2010-2012
Fiber-optic communication systems are nowadays facing serious challenges due to fast growing demand on capacity from various new applications and services.It is now well recognised that nonlinear effects limit the spectral efficiency transmission reach of modern fiber-optic communications.Nonlinearity compensation therefore widely believed be paramount importance for increasing future optical networks.Recently, there has been a steadily interest in application powerful mathematical tool -the...
Abstract Modern high-power lasers exhibit a rich diversity of nonlinear dynamics, often featuring nontrivial co-existence linear dispersive waves and coherent structures. While the classical Fourier method adequately describes extended waves, analysis time-localised and/or non-stationary signals call for more nuanced approaches. Yet, mathematical methods that can be used simultaneous characterisation localized fields are not yet well developed. Here, we demonstrate how Nonlinear transform...
We propose a method to improve the performance of nonlinear Fourier transform (NFT)-based optical transmission system by applying neural network post-processing spectrum at receiver. demonstrate through numerical modeling about one order magnitude bit error rate improvement and compare this with machine learning processing based on classification received symbols. The proposed approach also offers way accuracy inverse NFT; therefore, it can find range applications beyond communications.
Optical communication systems, operating in C-band, are reaching their theoretically achievable capacity limits. An attractive and economically viable solution to satisfy the future data rate demands is employ transmission across full low-loss spectrum encompassing O, E, S, C L band of single mode fibers (SMF). Utilizing all five bands offers a bandwidth up $\sim$53.5THz (365nm) with loss below 0.4dB/km. A key component realizing multi-band optical systems amplifier. Apart from having an...
We evaluate improvement in the performance of optical transmission systems operating with continuous nonlinear Fourier spectrum by artificial neural network equalisers installed at receiver end. propose here a novel equaliser designs based on bidirectional long short-term memory (BLSTM) gated recurrent and compare their several fully connected layers. The proposed approach accounts for correlations between different spectral components. application BLSTM leads to 16x terms bit-error rate...
Abstract The deployment of artificial neural networks-based optical channel equalizers on edge-computing devices is critically important for the next generation communication systems. However, this still a highly challenging problem, mainly due to computational complexity networks (NNs) required efficient equalization nonlinear channels with large dispersion-induced memory. To implement NN-based equalizer in hardware, substantial reduction needed, while we have keep an acceptable performance...
This roadmap consolidates recent advances while exploring emerging applications, reflecting the remarkable diversity of hardware platforms, neuromorphic concepts, and implementation philosophies reported in field. It emphasizes critical role cross-disciplinary collaboration this rapidly evolving
In this work we present a novel implementation of delay line free reservoir computing based on state-of-the-art photonic technologies, which exploits chaotic optical frequency comb formation in microresonator as the nonlinear reservoir. Our solution leverages high resonator Q-factor both for memory and enhancing dimensional mapping input symbols. We numerically demonstrate accurate prediction about one thousand symbols time series without need dedicated optimisation specific tasks. results...
In this work, we introduce the periodic nonlinear Fourier transform (PNFT) method as an alternative and efficacious tool for compensation of transmission effects in optical fiber links. Part I, algorithmic platform technique, describing details direct inverse PNFT operations, also known scattering (in time variable) Schrödinger equation (NLSE). We pay a special attention to explaining potential advantages PNFT-based processing over previously studied (NFT) based methods. Further, elucidate...
In this paper we propose the design of communication systems based on using periodic nonlinear Fourier transform (PNFT), following introduction method in Part I. We show that famous "eigenvalue communication" idea [A. Hasegawa and T. Nyu, J. Lightwave Technol. 11, 395 (1993)] can also be generalized for PNFT application: case, main spectrum attributed to signal decomposition remains constant with propagation down optical fiber link. Therefore, encoded data same way as soliton eigenvalues...
Nonlinear Fourier transform (NFT) based transmission technique relies on the integrability of nonlinear Schrödinger equation (NLSE). However, lossless NLSE is not directly applicable for description light evolution in fibre links with lumped amplification such as Erbium-doped amplifier (EDFA) because nonuniform loss and gain evolution. In this case, path-averaged model usually applied an approximation true including loss. inaccuracy path-average model, even though being small, can also...
Most of the nonlinear Fourier transform (NFT) based optical communication systems studied so far deal with burst mode operation that substantially reduce achievable spectral efficiency. The requirement emerges due to very nature commonly used version NFT processing method: it can process only rapidly decaying signals, requires zero-padding guard intervals for dispersion-induced channel memory, and does not allow one control time-domain occupation well. Some limitations drawbacks imposed by...
In this article, for the first time, a full-spectrum periodic nonlinear Fourier transform (NFT)-based communication system with inverse transformation at transmitter performed by using solution of Riemann-Hilbert problem (RHP), is proposed and studied. The entire control over spectrum rendered our technique, where we operate two qualitatively different components represented, correspondingly, in terms main phases, allows us to design time-domain signal tailored characteristics transmission...
We propose a modification of the nonlinear digital signal processing technique based on inverse synthesis for systems with distributed Raman amplification. The proposed path-average approach offers 3 dB performance gain, regardless power profile.
Abstract We combine the nonlinear Fourier transform (NFT) signal processing with machine learning methods for solving direct spectral problem associated Schrödinger equation. The latter is one of core science models emerging in a range applications. Our focus on unexplored computing continuous spectrum decaying profiles, using specially-structured deep neural network which we coined NFT-Net. Bayesian optimisation utilised to find optimal architecture. benefits NFT-Net as compared...
We apply both the unsupervised and supervised machine learning (ML) methods, in particular, k-means clustering support vector (SVM) to improve performance of optical communication system based on nonlinear Fourier transform (NFT). The NFT employs continuous spectrum part carry data up 1000 km using 16-QAM OFDM modulation. classify terms BER versus signal power dependence. show that can be improved considerably by means ML techniques more advanced SVM method typically outperforms clustering.
In this work we introduce the periodic nonlinear Fourier transform (PNFT) and propose a proof-of-concept communication system based on it by using simple waveform with known spectrum (NS). We study performance (addressing bit-error-rate (BER), as function of propagation distance) transmission use PNFT processing method show benefits latter approach. By analysing our simulation results for lumped amplification, demonstrate decent potential new method.
We evaluate, for the first time, achievable spectral efficiency of periodic nonlinear Fourier transform based communication systems with hard decision FEC and modulated perturbed plane waves high order QAM formats, e.g 32QAM-512QAM.
By performing the exact inverse transformation, a periodic solution to channel model is constructed and used in an NFT-based communication system. The achievable mutual information calculated using non-uniform probability distribution for transmitted symbols different link lengths.
We compare performance of several machine learning methods, including support vector machine, k-nearest neighbours, k-means clustering, and Gaussian mixture model, used for increasing transmission reach in the optical communication system based on periodic nonlinear Fourier transform signal processing.
We present a novel fibre-optic transmission system based on the phase modulation via nonlinear Fourier transform for finite-genus signals with no periodicity constraint, which is achieved through designing neural network-based receiver and demonstrate, as proof of concept, signalling below 7% HD-FEC BER threshold.
The authors propose a computationally efficient spectrum sensing solution for an orthogonal frequency division multiplexing (OFDM) signal in frequency-selective fading channel with additive white Gaussian noise (AWGN). authors' assumption is that the data symbols, coefficients and variance are all unknown. nature of problem leads them to find invariant detector. optimum one uniformly most powerful (UMPI); their effort shows this test does not exist, as final decision statistic depends on...
Summary form only given. Fibre lasers are known to provide a rich tapestry of operational regimes, which can be attributed the nonlinear nature light dynamics in optical fibre at high powers, and multidimensional system parameter space. Given their inherent complexity, identifying discerning underlying physical processes that gives rise them still remains formidable challenge. Here, for first time experiment, we show how Nonlinear Fourier Transform (NFT) (see e.g. [1-3] references therein)...