Strong and stable Data communication Using Artificial Intelligence method in Mobile Ad-Hoc Networks

Abstract

Using the idea of machine learning, the authors of this study provide a novel security mechanism for an Internet of Things model that is built on a mobile ad hoc network (MANET). The Black Hole Assault, also known as BHA, is among the most dangerous threats that can be found in a MANET. In this attack, the attacker node eliminates all of the data flow, which brings the efficiency of the model down. As a consequence of this, the development of a technique that is capable of shielding the system from the BHA node is required. This paper presented Ad-hoc On-Demand Distance Vector. It is a new advanced routing protocol which provides a combination of three different methods: ABC, ANN & SVM. The innovation of the developed framework is the integration of the SVM with the ANN, which allows detecting the attackers inside the identified path by using the AODV routing algorithm. This helps to prevent attacks. In this case, the ANN learning algorithm is used to train the model, while the ABC optimization algorithm and SVM are used to pick the training examples. The purpose of ABC is to improve the path that the transfer of information takes here between origin as well as the node at which it is ultimately received. The ABC algorithm's suggestion of the most efficient path, coupled with the attributes of each node, is then input into the SVMmodel. ANN determines if the node in question is a regular node or an intruder node by examining these qualities. The numerical simulation that was carried out in MATLAB demonstrates that the research that has been presented demonstrates an advantage in terms of Packet Delivery Ratio,latency, and throughput.