VigiNIDS: Network Intrusion Detection System

What is a Network Intrusion Detection System (NIDS)?

A Network Intrusion Detection System (NIDS) is a cybersecurity tool designed to monitor and analyze network traffic for malicious activities or policy violations. It plays a critical role in identifying potential threats in real-time and ensuring the security of systems and data.

Why is NIDS Important?

• Detects unauthorized access and cyberattacks.
• Helps organizations protect sensitive data and prevent breaches.
• Provides actionable insights to enhance network security.

---

Dataset Overview

Dataset Used: UNSW-NB15

The UNSW-NB15 dataset is a benchmark dataset for evaluating intrusion detection systems. It contains real-world network traffic data combined with synthetic attack traffic.

Key Details:

• Source: UNSW-NB15 Dataset
• Features: 49 features describing network traffic behavior.
• Records:
  • Training set: 175,341 rows.
  • Testing set: 82,332 rows.
• Classes:
  • 0: Normal traffic.
  • 1: Attack traffic.

---

Insights from the Data

Basic Observations:

• The dataset is well-structured and contains no missing values.
• There are only 4 categorical features that require encoding.
• In the train_df, there is a higher proportion of attack traffic compared to normal traffic. Conversely, the test_df has more normal traffic than attacks.

Correlation Analysis:

• Features like sbytes and spkts are highly correlated, indicating potential redundancy.
• Many features have weak correlations with the target label (label), suggesting they might contribute less to the classification task.

---

Visualizations So Far

Target Variable Distribution

• The distribution shows the proportion of Normal and Attack labels in the dataset.

Correlation Heatmap

• A heatmap showcasing relationships between features in the dataset. Strong correlations like those between sbytes and spkts are visible.

---

Next Steps

1. Perform data preprocessing:
   • Encode categorical features.
   • Scale numerical features if necessary.
2. Train and evaluate machine learning models.
3. Fine-tune the best-performing models.
4. Present final results and insights.

---

References

• UNSW-NB15 Dataset: Link to Dataset