Effective model for security against Vampire attacks in assorted wireless networks

Abstract

: In Wireless Sensor Networks, energy efficiency is crucial for intermediate sensor nodes. Due to the ad hoc organization of nodes, ensuring security in WSNs is a challenging task. Vampire attacks, which drain the battery life of sensor nodes, can paralyze the network. These attacks are difficult to detect and prevent, as they exploit the energy requirements for packet transmission from source to destination. Malicious nodes, or vampire packets, gradually weaken the energy of nodes, leading to network failure. Extending the lifespan of nodes necessitates the detection and avoidance of vampire packets. This thesis focuses on vampire attacks and their defenses at the routing protocol layer, proposing various techniques for their detection and control. newlineMobile ad hoc networks, with their decentralized, self organized, and open nature, are vulnerable to various attacks. Successful data transmission in these networks relies on node cooperation, which can be achieved using trust information. This thesis proposes a fuzzy logic-based trust model to mitigate the effects of vampire attacks. A single vampire attack can significantly deplete network resources, such as battery power. The performance of the proposed model is measured using precision, recall, and communication overhead. newlineThe thesis proposes a novel approach for vampire attack detection and prevention by predicting energy consumption in the data path with minimal error. Using a combination of the social spider optimized Gaussian mixture model and the grey prediction model, the scheme calculates a cooperative trust score to detect and prevent attacks. The algorithm, validated against recent state of the art schemes, achieves a detection accuracy improvement of up to 35.27 percent. newline newline