Vampire Attack Detection by Applying Support Vector Machine Algorithm in Wireless Sensor Network
B. Victoria Jancee, S. Kalaiselvi, K. Muthukumaran, K. Vijaya Bhaskar · 2023
Denial of Service (DoS) attacks at the routing levels were earlier security's primary focus. Resource depletion attacks were formerly exclusively thought of as routing issues, but they have lately been included in a new category as “vampire attacks.” Every protocol under consideration is vulnerable to vampire attacks, which are destructive, difficult to identify, and simple to execute, with only one rogue insider sending only protocol-compliant communications. To solve these problems, Vampire attack detection by applying a Support Vector Machine (SVM) algorithm in a wireless sensor network (VSVM) is proposed. This mechanism uses the SVM algorithm to separate the vampire nodes ideally. The SVM algorithm isolates the vampire nodes based on sensor node characteristics like node energy, node broadcast count, and node packet received ratio (PRR). Simulation results reveal that the VSVM mechanism enhances the vampire attack detection ratio and throughput. Moreover, it minimizes the delay and routing overhead in the wireless sensor network.