vioft2nntf2t|tblJournal|Abstract_paper|0xf4ff28f42b00000081700a0001000600 The virtual and physical worlds are bridged using the largest digital mega-trend called the Internet of Things (IoT). Between mankind, new interactions and new business models are emerging due to the incremental growth in the Internet, machines, objects, and people connectivity. Secured communication is a typical challenge that is raised due to IoT high diversity, restricted computational resources, and protocols and standards. Because of the huge attack surface in IoT networks, they are highly vulnerable to various attacks, even with some security measures. So, for detecting attacks, it is necessary to design defense mechanisms. In IoT environments, it is highly crucial to have security defense measures like Intrusion Detection Systems (IDS). Hence, authentication and encryption traditional security countermeasures are not sufficient. At network level, to solve those issues and to protect Internet-connected frameworks, major solutions are provided by IDS. Highly unique challenges are faced by IoT specific characteristics like malware detection, ransomware, processor architecture heterogeneity, and the gap in security design. However, as in literature, various problems are raised in traditional IDS, like the high false alarm rate. In IoT, for intrusion detection, a detailed study of traditional Deep Learning (DL) and Machine Learning (ML) techniques and recent technologies is presented in this review. For presenting every selected work objective and methodology, they are analysed and this review work discusses their results. IoT systems cannot be secured by applying traditional security techniques directly due to their computational constraints and intrinsic resources. In real time, on IoT devices, unknown and known attacks are detected using ML techniques in IDS. An IDS is presented in this review and its working is independent of network structure and IoT protocols. This IDS do not require any prior knowledge of security threats. Therefore, for providing security as a service to IoT networks, an artificially intelligent IDS is developed. This review paper provides a clear discussion of various attack detection techniques, along with their benefits and drawbacks.
R Anushiya1, V S Lavanya2
P.K.R. Arts College for Women, India1, P.K.R. Arts College for Women, India2
Genetic Algorithms (GA), Deep Learning (DL), Intrusion Detection Systems (IDS), Internet of Things (IoT)
| January | February | March | April | May | June | July | August | September | October | November | December |
| 1 | 0 | 0 | 3 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 |
| Published By :
ICTACT
Published In :
ICTACT Journal on Communication Technology
( Volume: 12 , Issue: 3 , Pages: 2527-2537 )
Date of Publication :
September 2021
Page Views :
136
Full Text Views :
10
|