Hybrid Intrusion Detection Technique for Internet of Things

Title

Hybrid Intrusion Detection Technique for Internet of Things

Creator

Jinsi, Jose

Contributor

Deepa, V Jose.

Description

The rapid expansion and integration of Internet of Things (IoT) applications in newlinevarious aspects of daily life has significantly surprised and impacted contemporary society. The most crucial keyword concerning these applications is security, specifically, in the enormous amount of data generated every second, and how it is used. These applications are vulnerable to various attacks, which could result in an unthinkable catastrophe if not managed and controlled with sufficient foresight. Growing concerns about data security in the expanding IoT landscape are driven by factors such as increased vulnerability of devices to viruses, susceptibility to denial-of-service attacks, and heightened risk of intrusion attempts. To prevent such occurrences, stronger precautions should be taken, enabling system developers and manufacturers of IoT devices to enhance their approaches to better security mitigation. It is essential to identify all potential threats and vulnerabilities that are created explicitly for IoT infrastructures. It is believed that to lessen potential dangers, there is a need for more significant research on security attacks. Security difficulties have been found and must be dealt with, so they may be avoided. Further research must address security challenges in IoT-based environments, particularly for suppliers and consumers, to gradually raise the reliability of IoT applications. Although many conventional methods are still used, there might be superior options for devices with limited resources. Artificial intelligence plays a significant role in this issue. newlineThis research first tries to comprehend how machine learning methods relate to attack newlinedetection. The effects of different machine learning techniques are evaluated using the newlineUNSW-NB 15 dataset. Additionally, it has been found that each model performs worse overall, mainly when security issues are present. As a result, real-time datasets and Deep Learning (DL) algorithms for intrusion detection in the IoT need to be prioritized.

Source

Author's Submission

Date

2023-01-01

Publisher

Christ(Deemed to be University)

Subject

Computer Science

Rights

Open Access

Relation

61000286

Format

PDF

Language

English

Type

PhD

Identifier

http://hdl.handle.net/10603/541888