Proposed machine learning system to predict and estimate impulse noise in OFDM communication system

Ali N. Hasan, Thokozani Shongwe

Research output: Chapter in Book/Report/Conference proceedingConference contributionpeer-review

8 Citations (Scopus)

Abstract

This paper investigates the use of machine learning (ML) in predicting and estimating the impulse noise. Four ML's algorithms (Multilayer perceptron MLP, support vector machine SVM, k nearest neighbour kNN and naïve Bayesian classifier NBC) were implemented in an OFDM system affected by impulse noise. The impulse noise model used was the Middleton Class A noise model. The ML's were trained with Middleton Class A impulse noise model so that they can be able to predict the presence of impulse noise in the communication system. In terms of prediction accuracy, results showed that kNN slightly outperformed MLP and NBC and accomplished high prediction accuracy of 99.8%. SVM achieved the lowest prediction accuracy among the four used methods. These results indicates that machine learning could be used to estimate impulse noise in OFDM communications system.

Original languageEnglish
Title of host publicationProceedings of the IECON 2016 - 42nd Annual Conference of the Industrial Electronics Society
PublisherIEEE Computer Society
Pages1016-1020
Number of pages5
ISBN (Electronic)9781509034741
DOIs
Publication statusPublished - 21 Dec 2016
Event42nd Conference of the Industrial Electronics Society, IECON 2016 - Florence, Italy
Duration: 24 Oct 201627 Oct 2016

Publication series

NameIECON Proceedings (Industrial Electronics Conference)

Conference

Conference42nd Conference of the Industrial Electronics Society, IECON 2016
Country/TerritoryItaly
CityFlorence
Period24/10/1627/10/16

Keywords

  • Impulse noise
  • Multilayer perceptron
  • Naïve Bayesian
  • OFDM
  • Prediction
  • Support vector machines

ASJC Scopus subject areas

  • Control and Systems Engineering
  • Electrical and Electronic Engineering

Fingerprint

Dive into the research topics of 'Proposed machine learning system to predict and estimate impulse noise in OFDM communication system'. Together they form a unique fingerprint.

Cite this