Distributed Denial of Service (DDoS) Detection Tool Using Artificial Neural Network (ANN) for Homelab
Keywords:
Homelab, Intrusion Detection System, Artificial Intelligence, Distributed Denial of Service, Artificial Neural Networks, Raspberry PiAbstract
A homelab is a personal server or cluster of servers kept in a residence and used for various purposes such as testing, development, or everyday use. These labs enable tech enthusiasts to create their own video streaming services, email hosting platforms, cloud storage, and other services without relying on large tech corporations. However, the increasing presence of advanced botnets and the search for vulnerable databases to steal and sell data from has posed a threat to many homelab users who do not have the budget or resources to implement a comprehensive intrusion detection system (IDS). To address these challenges, a proposal has been made to create a Distributed Denial of Service (DDoS) detection tool using Artificial Neural Networks (ANN) for homelab users. The goals of this proposed system are to design and develop an ANN-based DDoS detection tool for homelab users, and to test its effectiveness against real-life threats. The expected outcome is the implementation of an ANN capable of accurately detecting DDoS attacks with minimal false positives without hindering the performance of the server. Currently, the system currently has a recall and F1 score of 0.97 which indicates that the system is able to detect DDoS attack with fewer false positives.