"

Cookies ussage consent

I agree.

Our site saves small pieces of text information (cookies) on your device in order to deliver better content and for statistical purposes. You can disable the usage of cookies by changing the settings of your browser. By browsing our site without changing the browser settings you grant us permission to store that information on your device.

DWDM reconstruction using supervised and unsupervised learning approaches

K. VENKATESAN1,* , A. CHANDRASEKAR2, P. G. V. RAMESH3

Affiliation

  1. Department of ECE, St. Joseph’s College of Engineering, Chennai, India
  2. Department of CSE, St. Joseph’s College of Engineering, Chennai, India
  3. Department of ECE, St. Joseph’s Institute of Technology, Chennai, India

Abstract

This paper attempts to reconstruct and optimize a DWDM system by avoiding distortions such as Four-Wave Mixing (FWM) and high signal distortion due Inter-Channel Interference (ICI) using supervised and unsupervised learning approaches. FWM in high channel DWDM system reduces the network flexibility, transmission capacity and increases computational difficulties. The ICI and Signal distortion that affects spectral efficiency, increases network latency, leading to undesirable data re-transmission. To solve these above stated issues, the DWDM system necessitates optical variables optimization using supervised and unsupervised regression learning. In this paper, we reconstruct the DWDM system design using supervised and unsupervised regression learning approaches, which are used to identify, correlate, and optimize the FWM influencing optical parameters. Furthermore, trained datasets are generated from parameter-based simulations. Results are analyzed using supervised and unsupervised regression approaches, which improves the DWDM mechanism and achieves accuracy through a computerized regression model controller. Thus, the reconstructed DWDM system is re - designed with optimized FWM parameters obtained through supervised machine learning approaches, and unsupervised training evaluates the proposed R-DWDM system to predict Q-factor, OSNR, signal, and noise power levels accurately.

Keywords

Supervised and unsupervised approaches, DWDM, Optical parameters, Regression model controller, Correlation, FWM.

Citation

K. VENKATESAN, A. CHANDRASEKAR, P. G. V. RAMESH, DWDM reconstruction using supervised and unsupervised learning approaches, Optoelectronics and Advanced Materials - Rapid Communications, 15, 9-10, September-October 2021, pp.459-470 (2021).

Submitted at: Oct. 26, 2020

Accepted at: Oct. 7, 2021