Evaluation of various ML (Machine Learning) algorithms to detect attacks in Mobile Ad hoc Networks (MANETs) via AODV

Infrastructure-less networks pose significant challenges in data security and time management, particularly within Mobile Ad-Hoc Networks (MANETs), where unstable mobile nodes can disrupt routing and affect Quality of Service (QoS) metrics.While numerous solutions exist for addressing these challenges, many of them introduce increased Overhead (OH) and Normalized Routing Load (NRL) to the MANET, which is unacceptable given the time-sensitive nature of communication sessions in MANETs.Exceeding processing or transmission times can lead to issues like Link Break (LB).The author proposes a solution to these challenges within the IDS-ATiC-AODV framework (Improved Data Security -Avoiding Time Complexity Ad-Hoc On-Demand Distance Vector).This IDS-ATiC AODV framework addresses five distinct qualitative and quantitative QoS issues using two primary algorithms.This study evaluates various Machine Learning algorithms from different clusters, leveraging the Infrastructure-Less Knowledge Measure Source Dataset (Infra-Less KMS Dataset) for analysis.This involves assessing the accuracy of each Machine Learning algorithm across ten different challenges using the Infra-Less KMS Dataset.