Wormhole Attack Detection Technques In MANET

Abstract

MANET is a wireless medium, with no infrastructure, for nodes to communicate within a network. Movement of nodes is a lot more in this network than other networks due to lack of infrastructure (Wireless nature). In a mobile ad hoc network, each node functions as a router and finds the best path forwarding a packet via origin to destination. This network is important in places where building a framework is impossible, like in a military setting. A mobile ad-hoc network (MANET) is a temporary network created by unrestricted nodes that can move around and communicate without a centralized network. In a black hole attack, several or even more attacker node use a tunnel-like private channel to find a wormhole attack. Wormhole tunnel will then start collecting the data packets and transferring them to another location. The proposed work encouraged secure and effective routing in MANET. At that point, we selected 20 important attributes to create a dataset that was labelled with the aid of a unique node address. Apply two well-known machine learning classifiers that categorise data into two groups, specifically normal and malicious, to test sample data in this manner. For feature selection and SVM (Support Vector Machine) classification, respectively, we employ genetic algorithms. The demonstration of the technology was assessed on many quantifiable characteristics and compared to contemporary techniques.