Deep learning based network attack classification
Year: 2025
Type: article
Abstract: The most common security threads to internet service is distributed denial of service also known as DDoS attack. This DDoS can be easily be launched in pc,cloud service providers and many more that suddenly spikes in traffic or websites slowdown. There are traditional DDoS attack detection are present which uses ML based algorithms gives accuracy but if the dataset consists of large volume of data the traditional algorithms find difficult to detect. Since Deep learning is widely used in present days, which is powerful learning and has extraction capabilities. In this paper we used LSTM which is a type of Recurrent Neural Network (RNN) which are Deep Learning based algorithms. In this paper WSN-DS.cv dataset is used, the data set consists of wireless sensor network classifications. Here we train the raw data by preprocessing of data like feature extraction, identifying the null values, analysing the data and splitting them for train and test. Here we epoch the dataset for 10 times to get better accuracy
Authors: Nagarjuna Reddy.G -, Dinesh.V -, Pavan Kumar.K -, Jeeva.V -, Kiruba Devi.T -
Language: English
Subfield: Computer Networks and Communications
Field: Computer Science
Domain: Physical Sciences
Open Access status: hybrid
OpenAlex ID: W4408909168