Elsevier

Optics Communications

Volume 470, 1 September 2020, 125819
Optics Communications

A modulation format identification method based signal amplitude sorting and ratio calculation

https://doi.org/10.1016/j.optcom.2020.125819Get rights and content

Highlights

  • A MFI method based on signal amplitude sorting and ratio calculation is proposed.

  • The method is insensitive to phase impairments.

  • The proposed MFI method can achieve accurate identification at different impairments.

  • The method can be deployed in the practical OPM system without additional hardware.

Abstract

Modulation format identification (MFI) is an indispensable part of next generation optical networks such as elastic optical networks (EON). We propose a simple MFI method based on signal amplitude sorting and ratio calculation. After sorting the amplitude of signals, three MFI feature values are extracted to identify polarization division multiplexing (PDM) quadrature phase shift keying (PDM-QPSK) and three commonly used quadrature amplitude modulation (QAM) signals (PDM-16QAM, PDM-32QAM and PDM-64QAM). Numerical simulations verify the effectiveness of proposed MFI method with the required optical signal noise ratio (OSNR) lower than 7% forward error correction (FEC) threshold. Proof-of-concept experiments using QPSK, 16QAM, 32QAM and 64QAM signals with 100 km fiber transmission demonstrate the feasibility of proposed MFI method. We also investigate the influence of fiber transmission distance, fiber nonlinearity and symbol numbers on the proposed method and the results show that the proposed MFI methods can achieve accurate identification at different fiber transmission length, a range of fiber nonlinearity and small requirement of symbol number. The method is simple and insensitive to phase impairments caused by laser phase noise and frequency offset (FO), which will bring convenience into future EONs.

Introduction

To satisfy the demands for ever-growing capacity requirement of big data, cloud computing and streaming video, optical fiber communication systems have been significantly improved toward large capacity, long transmission distance, and high spectral efficiency [1], [2], [3], [4]. The next generation flexible and cognitive optical networks have attracted worldwide research interests and become more dynamic and heterogeneous [5], [6].

Elastic optical networks (EONs) based on orthogonal frequency division multiplexing (OFDM) technology has been proposed over the past few years [7]. To maximize the bandwidth and energy efficiency, different transmission parameters like modulation formats (MFs), symbol rates of data, optical launch power etc. can be adjusted adaptively in EON based on time-varying channel conditions and traffic demands [8]. Modulation format identification (MFI) is indispensable in the optical receivers because of the dynamic variation of transmission parameters. For example, the carrier recovery module and phase noise compensation must be appropriate to the received MF.

Various MFI techniques have been put forward recently. A. Swami and B. M. Sadler proposed a MFI method based on fourth-order cumulant features after correcting phase errors [9]. S. M. Bilal et al. proposed a peak-to-average-power ratio based MFI scheme which requires optical-signal-to-noise ratio (OSNR) as prior information [10]. G. Liu et al. proposed a blind MFI method by using nonlinear power transformation and peak detection. The method has strong tolerance amplifier spontaneous emission (ASE) noise, but cannot achieve higher-order MFI such as quadrature amplitude modulation (QAM) signals [11]. The Stokes-space-based MFI methods have insensitivity to carrier phase noise and frequency offset (FO) and has attracted many researchers. These MFI methods must be performed after tracking the state of polarization and recovering the initial polarization state [12], [13], [14], [15], [16]. Nowadays, various machine learning methods are introduced in MFI based on the concept of distribution of data such as artificial neural network, deep neural network and convolutional neural network. To construct the neural network, a large number of training data and complicated training processing are needed [17], [18], [19], [20], [21], [22]. X. Lin et al. proposed a joint modulation format identification and OSNR estimation algorithm which relies on the cumulative distribution function (CDF) of the received signal’s amplitude in combination with support vector machine (SVM)-based classification and regression for coherent optical receivers. The SVM-based method shows good classification accuracies and small mean estimation error [23]. There need 10000 symbols for creating the CDF. Y. Zhao et al. proposed a low-complexity and joint MFI and OSNR estimation scheme using random forest which needs 8250 training data [24].

In this work, we propose a MFI method based on signal amplitude sorting and ratio calculation. In Section 2, we introduce the principle of the proposed MFI method. After sorting the amplitude of signals, MFI feature values are extracted to identify the polarization division multiplexing quadrature phase shift keying (PDM-QPSK), PDM-16QAM, PDM-32QAM and PDM-64QAM signals. In Sections 3 Simulation results, 4 Results of proof-of-concept experiments, numerical simulations verify the effectiveness of proposed MFI method with the required OSNR lower than 7% FEC threshold. Proof-of-concept experiments using QPSK, 16QAM, 32QAM and 64QAM signals with 100 km fiber transmission also confirm the feasibility of the proposed MFI method. We also investigate the influence of fiber transmission distance, fiber nonlinearity and symbol numbers on the proposed method and the results demonstrate that the proposed MFI methods achieve accurate identification at different fiber transmission, a range of fiber nonlinearity and small requirement of symbol number. In Section 5, we conclude that, the method is simple and insensitive to phase impairments caused by laser phase noise and FO, which will bring convenience into future EONs.

Section snippets

Principle

Here we define some basic notions of the proposed MFI method. For an unknown MF signal data, the amplitudes of symbols in ascending order are defined as a1a2aN, where N is symbol number. We get set Asort=a1,a2,,aN and define subset AsortN1N2=aN1,aN1+1,,aN2 where 1N1<N2N. In addition, the elements in the subset would not be merged if two or more elements are identical. For example, for ideal 16QAM signal with 1000 symbols, the Asort can be categorized into three subset: Asort10.25N, Aso

Simulation results

A simulation system is set up by commercial software Virtual Photonics Inc. Transmission Maker to verify the proposed MFI method. The schematic diagram of the proposed method used in a coherent optical fiber transmission system is shown in Fig. 2. At the transmitter, the center wavelength of the laser is set at 1553.6 nm and the linewidth is 100 kHz. Two pseudo-random binary sequences (PRBS) are produced by an arbitrary waveform generator (AWG) and drive I and Q branches independently. The

Results of proof-of-concept experiments

We built a single-polarization fiber transmission system with QPSK, 16QAM, 32QAM and 64QAM signals. The available experimental configuration is as following: At the transmitter, the center wavelength is 1550 nm with the linewidth of 5 kHz. AWG generates two independent PRBS and drive I and Q branches of IQ modulator. 10 GBaud modulated optical signal is transmitted in the fiber link with 0 dBm launch power. The fiber link is constructed by 100 km SSMF and EDFA. The parameters of SSMF are shown

Conclusions

In this work, a MFI scheme based on signal amplitude sorting and ratio calculation is proposed. After sorting the amplitude of signals, three MFI feature values R1,R2,R3 are extracted to identify the PDM-QPSK, PDM-16QAM, PDM-32QAM and PDM-64QAM signals. 28 GBaud PDM-QPSK, PDM-16QAM, PDM-32QAM and PDM-64QAM signals with 2000 km fiber transmission distance are researched to verify the proposed method. The minimum required OSNR values for 100% successful identification of four MF signals are 5 dB,

Declaration of Competing Interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Acknowledgments

This research was funded by National Natural Science Foundation of China, grant number 61427813 and Open Fund of IPOC (BUPT) , grant number IPOC2018B003.

References (28)

  • ZhaoY. et al.

    Low-complexity and joint modulation format identification and osnr estimation using random forest for flexible coherent receivers

    Opt. Commun.

    (2020)
  • GerstelO. et al.

    Elastic optical networking: a new dawn for the optical layer

    IEEE Commun. Mag.

    (2012)
  • XiaT.J. et al.

    Introduction of spectrally and spatially flexible optical networks

    IEEE Commun. Mag.

    (2015)
  • ColeChris

    Beyond 100G client optics

    IEEE Commun. Mag.

    (2012)
  • YanL.-S. et al.

    Toward the Shannon limit of spectral efficiency

    IEEE Photonics J.

    (2011)
  • NapoliA. et al.

    Next generation elastic optical networks: The vision of the european research project IDEALIST

    IEEE Commun. Mag.

    (2015)
  • LuoY. et al.

    Time-and wavelength-division multiplexed passive optical network (TWDM-PON) for next-generation PON stage 2 (NG-PON2)

    J. Lightwave Technol.

    (2012)
  • ZhaoZ. et al.

    A modulation format identification method based on amplitude deviation analysis of received optical communication signal

    IEEE Photonics J.

    (2015)
  • TomkosI. et al.

    A tutorial on the flexible optical networking paradigm: State of the art, trends, and research challenges

    Proc. IEEE

    (2014)
  • SwamiA. et al.

    Hierarchical digital modulation classification using cumulants

    IEEE Trans. Commun.

    (2000)
  • BilalS.M. et al.

    Blind modulation format identification for digital coherent receivers

    Opt. Express

    (2015)
  • LiuG. et al.

    Blind modulation format identification using nonlinear power transformation

    Opt. Express

    (2017)
  • BoT. et al.

    Modulation format recognition for optical signals using connected component analysis

    IEEE Photonics Technol. Lett.

    (2016)
  • MaiX. et al.

    Stokes space modulation format classification based on non-iterative clustering algorithm for coherent optical receivers

    Opt. Express

    (2017)
  • Cited by (0)

    View full text