DDoS Attack Detection Using Long Short-Term Memory Based on Hybrid Grey Wolf Optimization and Tabu Search
Author(s)
Abstract
A Distributed Denial of Service (DDoS) attack in a Hadoop environment can have serious consequences, as it can disrupt the availability and performance of critical services and applications that rely on the Hadoop infrastructure for data storage and processing. An intriguing technique that makes use of deep learning capabilities to examine network traffic patterns and spot unusual behavior suggestive of an attack is the use of Long Short-Term Memory (LSTM) networks for DDoS attack detection in a Hadoop context. However, the LSTM algorithm has several shortcomings such as low accuracy, and convergence rate due to its improper selection of hyperparameters. Hence, the optimized LSTM method is proposed based on hybrid grey wolf optimization (GWO) and tabu search (TS) called GWO-TS. The hybrid GWO-TS method is used to optimize the hyperparameters of LSTM which is applied to detect the DDoS attack in the Hadoop environment. The experimental results show that the developed optimized LSTM produced high detection accuracy with fast convergence when compared to literature algorithms.
Keywords