DDoS Detection and Defense Based on PSO Optimized LSTM Neural Network
Yaowen Dou, Zhenliang Li, Junhao Li, Yu Lu · 2023
In recent years, DDoS attacks targeting iot devices have become more and more frequent. Identifying and defending DDoS attacks efficiently, correctly and quickly has become a hot issue at present, which has important practical application value. In this paper, a LSTM neural network model based on particle swarm optimization (PSO) is designed. The LSTM neural network model extracts the features of various messages and data packets when the system encounters DDoS attacks, and uses the PSO optimization algorithm to optimize the parameters globally, and finally finds the optimal hyperparameters of the current model, so as to improve the detection and defense ability of the system against DDoS attacks. Through experiments, the PSO-LSTM model proposed in this paper is higher than the comparison model in accuracy, recall rate and F1 value. The experimental results also show that compared with the traditional neural network, the LSTM neural network model optimized by the PSO algorithm has a significant improvement in the performance of DDoS attack detection.