Intensive analysis of intrusion detection methodology over Mobile Adhoc Network using machine learning strategies

Mobile Adhoc Network (MANET) recently gained prominence due to the prevalence of handheld connectivity and their flexibility to supportability in specific non-permanent and instantaneous applications like floods and military situations. MANET offers great network utility, but comes with specific security challenges due to the fact that there is no central control, changing network topology, transient existence and uncoordinated communication. There are numerous proposals to use encryption and authentication measures to decrease the risk of security issues, especially as a first-line protection options. Although these risks cannot be removed entirely, an effective intrusion detection scheme is vital to keep unauthorized intrusion out of Mobile Adhoc Network. The role of intrusion identification on Mobile Adhoc Network is extremely difficult due to open medium, complex topology, dispersion, lack of centralized administration, and resource-constrained node groups. There is no direct analogue of a traditional intrusion detection system designed for Mobile AdHoc Networks Technology that can be used on the wireless network. The technology used in it must be flexible enough to accommodate ad hoc changes. This system implements new machine learning architecture that enhances detection to be much more. Intelligent Decision Support incorporates the high accuracy of Enhanced Support Vector Machine (eSVM) with the improved scalability of Rough Set Theory (RST). (c) 2021 Elsevier Ltd. All rights reserved. Selection and peer-review under responsibility of the scientific committee of the 1st International Conference on Computations in Materials and Applied Engineering - 2021.