Security method for internet of things using machine learning against cyber attacks

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

Abstract

Internet of things is a huge network of large number of devices with sensors. This group of devices is estimated to grow over 25 billion in 2020 as its growth since its evolution has been truly more than just rapid growth. With such a growth in network expansion and increase in number of connected devices, security always have been an issue to be improved and one of the weak areas in IoT environment. While applying security in IoT environment the main characteristics of IoT being heterogenetic and the number of IoT devices are the challenges that need to be dealt with some efficient and feasible approach. Therefore, to address this problem, we propose a method of securing IoT system using basic principles of machine learning. The system will analyze the data and detect the malicious packets received from the edge devices. The experimental study showed improved results with the dummy data sets in an IoT environment.

Original languageEnglish
Title of host publicationProceedings of the 2020 4th International Symposium on Computer Science and Intelligent Control, ISCSIC 2020
PublisherAssociation for Computing Machinery
ISBN (Electronic)9781450388894
DOIs
StatePublished - 17 Nov 2020
Event4th International Symposium on Computer Science and Intelligent Control, ISCSIC 2020 - Virtual, Online, United Kingdom
Duration: 17 Nov 202019 Nov 2020

Publication series

NameACM International Conference Proceeding Series

Conference

Conference4th International Symposium on Computer Science and Intelligent Control, ISCSIC 2020
Country/TerritoryUnited Kingdom
CityVirtual, Online
Period17/11/2019/11/20

Keywords

  • Artificial Intelligence
  • Cyber Security
  • Internet of Things

Fingerprint

Dive into the research topics of 'Security method for internet of things using machine learning against cyber attacks'. Together they form a unique fingerprint.

Cite this