PhD in Optical Communication Using Machine Learning

Introduction

In today’s time, there is an enormous amount of data in the cyber world. Optical network communication using machine learning is a technique that reduces the complexity of the architecture by using software Defined Network or simply SDN. SDN has the ability to for automation of network operations, including monitoring and troubleshooting. Moreover, it provides features like plug-n-play or zero-touch network operations. One of the most common usages of machine learning is for solving the classical Routing and Wavelength Assignment from previous data configuration.

Research Approach

In the proposed research of Optical network communication using machine learning, the foremost step is to gather good quality data and to create a code to understand the relation between the data variables.

Next up is building the architecture of the model by dividing it into various sets so as to prepare it for the process of fine tuning to optimally adjust its hyper-parameters which will lead to avoidance of over fitting and under fitting. The Deep Neural Network has been chosen for Optical network communication using machine learning that comprises of six linked layers, that consists of dropout in the first and fourth layer. This is also structured with 12 regularizations in the third and fifth layers.

Issues in the current technology

In the current technology, there is a Routing and wavelength Assignment problem and fiber access delay modeling. The proposed research on Optical network communication using machine learning is aimed to resolve both of these major shortcomings.