Multi-layer Perceptron Model for Mitigating Distributed Denial of Service Flood Attack in Internet Kiosk Based Electronic Voting

Abstract

Distributed Denial-of-Service (DDoS) flood attack targeting an Internet Kiosk voting environment can deprive voters from casting their ballots in a timely manner. The goal of the DDoS flood attack is to make voting server unavailable to voters during election process. In this paper, we present a Multilayer Perceptron (MLP) algorithm to mitigate DDoS flood attack in an e-voting environment and prevent such attack from disrupting availability of the vulnerable voting server. The developed intelligent DDoS flood mitigation model based on MLP Technique was simulated in MATLAB R2017a. The mitigation model was evaluated using server utilization performance metrics in e-voting. The results after the introduction of the developed mitigation model into the DDoS attack model reduced the server utilization from 1 to 0.4 indicating normal traffic. MLP showed an accuracy of 95% in mitigating DDoS flood attacks providing availability of voting server resources for convenient and timely casting of ballots as well as provide for credible delivery of electronic democratic decision making