Client Server Network Intrusion Detection System

Student: Aiyu Ahmad Musa (Project, 2025)
Department of Cyber Security
Nigerian Army University, Biu, Borno State

Abstract

The project focuses on building and evaluating a Network Intrusion Detection System (NIDS) using machine learning to detect malicious and normal network traffic. Using the NSL-KDD dataset, five models—Naïve Bayes, Decision Tree, K-Nearest Neighbors, Logistic Regression, and Artificial Neural Network (ANN)—were tested. The Decision Tree performed best with 87.5% accuracy, while Logistic Regression followed with 85.5%. ANN achieved 83%, showing potential for improvement. The study highlights how data preprocessing and model optimization improve accuracy but notes limitations like dataset imbalance and lack of real-time testing. It recommends future research on ensemble learning, deep learning optimization, and real-time implementation to strengthen cybersecurity.

Keywords
Network Intrusion Detection Client-Server Security Machine Learning Cybersecurity Artificial Neural Network Decision Tree NSL-KDD Dataset Intrusion Detection System Anomaly Detection Network Traffic Analysis
Filters
Institutions