An SVM-Based Detection for Coherent Optical APSK Systems With Nonlinear Phase Noise

Abstract

A support vector machine (SVM)-based data detection is proposed for coherent optical fiber amplitude phase-shift keying (APSK) communication systems where the nonlinear phase noise is the main system impairment. The performances of the detection with SVMs are investigated for three different 16-APSK modulation formats. In addition, three transmission scenarios with dispersion being considered or not are adopted to simulate and analyze the performances. Compared with the traditional two-stage maximum-likelihood detection, the SVM conducts detection without the need to know the information of transmission link, and it gains a relatively large improvement in the nonlinear system tolerance, particularly in the high nonlinear regime. Compared with quadrature amplitude modulation (QAM), the 16-APSK system can increase the nonlinear system tolerance by 4.88 dB at BER ¼ 1E À 3.Index Terms: Support vector machine (SVM), amplitude phase-shift keying (APSK), nonlinear phase noise (NLPN).

Machine Learning Techniques to Mitigate Nonlinear Phase Noise in Moderate Baud Rate Optical Communication Systems

“…Han et al [11] definition of nonlinear boundaries for demodulation, avoiding additional stages for phase estimation and correction, extraction of features from constellations of real transmission scenarios, demonstrating the potential of artificial intelligence algorithms in next-generation optical-wireless technologies. In [42], a description of the mathematical background of machine learning techniques from a signal processing perspective applied to nonlinear transmission systems, OPM, and cross-layer…”

“…The Voronoi diagram by contours has been used to assist the clustering-based demodulation stage when symbol distortion is severe [11]. An approximation to its definition may be introduced as the proximity regions to a set of objects in a bidimensional plane.…”

“…This same distortion appeared in our experimental scenario, corresponding to 16-Quadrature-Amplitude Modulation (QAM) and 4 + 12 Phase Shift Keying (PSK) constellations. To overcome PN impairments, well-known post-equalization techniques have been proposed, such as (i) detection methods based on the symbol error probability (SEP) in which minimum Euclidean distance (ED) decisions are used to define decision regions (for example, the rectangular grid for 16-QAM); (ii) demodulation methods based on maximum likelihood criterion applied for decision-making rules under conditions of uncertainty [10]; (iii) machine learning techniques applied to clusters, to classify and demodulate data symbol points, and mainly to improve the error system tolerance against nonlinear effects [11,12].…”

Machine Learning for Fiber Nonlinearity Mitigation in Long-Haul Coherent Optical Transmission Systems : Invited Paper

“…Due to its robustness, SVM is believed to separate received symbols to correct classifications well during nonlinearity equalization process. An SVM classifier was first trained and applied in nonlinearity compensation for combating nonlinear phase noise in amplitude phase-shift keying (APSK) system [14]. Recently, five SVM methods including: 1) the one versus rest (OvR) where the multi-classifiers are built one by one considering the rest belonging to the other class with the concept of binary SVM; 2) the symbol encoding; 3) the binary encoding (BE) is based on whether each bit of label feature is 0 or 1; 4) the constellation rows and columns (RC); and 5) the in-phase and quadrature components (IQC) were investigated and IQC indicates the optimal results among all five in terms of computing resource and hardware storage [15].…”

A Neuromorphic Silicon Photonics Nonlinear Equalizer For Optical Communications With Intensity Modulation and Direct Detection

“…Another set of digital approaches for nonlinear compensation is based on machine learning techniques. Some popular implementations are based on Artificial Neural Networks (ANNs) ( [7]) and Support Vector Machines (SVMs) ( [8], [9]). In [10] techniques based on nonlinear state-space based Bayesian filtering and Gaussian Mixture Models (GMMs) are presented.…”