Hybrid Intrusion Detection System untuk Keamanan Siber pada Jaringan Industri

Danendra Azriel Ramdhany, Drs. Bambang Nurcahyo Prastowo, M.Sc.

2024 | Skripsi | ELEKTRONIKA DAN INSTRUMENTASI

Perkembangan teknologi jaringan komputer dan penggunaan internet telah mengubah berbagai aspek kehidupan, termasuk sektor industri. Internet memungkinkan kontrol jarak jauh dalam proses produksi, tetapi juga meningkatkan risiko keamanan siber pada Industrial Control System (ICS) dan Supervisory Control and Data Acquisition (SCADA). Penelitian ini mengkaji ancaman terhadap ICS/SCADA dan mengusulkan penggunaan Intrusion Detection System (IDS).

Penelitian ini mengimplementasikan hybrid intrusion detection system dengan menggabungkan metode anomaly-based detection berbasis machine learning algoritme random forest dan signature-based detection menggunakan Suricata. Hasil analisis dari sistem divisualisasikan dalam dashboard serta dilakukan tindakan saat terjadi serangan.

Hasil dari penelitian ini menunjukkan bahwa metode hybrid efektif dalam mendeteksi serangan dan mengatasi kekurangan masing-masing metode. Proses deteksi divisualisasikan dengan baik melalui dashboard, memudahkan pemantauan secara real-time. Selain itu, sistem berhasil memblokir alamat IP penyerang yang menunjukkan potensi implementasi untuk meningkatkan keamanan siber di sektor industri.

The advancement of computer network technology and internet usage has transformed various aspects of life, including the industrial sector. The internet enables remote control in production processes but also heightens cybersecurity risks within Industrial Control Systems (ICS) and Supervisory Control and Data Acquisition (SCADA) systems. This research examines threats to ICS/SCADA and proposes the implementation of an Intrusion Detection System (IDS).

This study implements a hybrid intrusion detection system by combining anomaly-based detection using the random forest machine learning algorithm and signature-based detection through Suricata. The analysis results from the system are visualized on a dashboard, and actions are taken when an attack occurs.

The results of this study demonstrate that the hybrid method is effective in detecting attacks and addressing the limitations of each individual method. The detection process is well-visualized through a dashboard, facilitating real-time monitoring. Furthermore, the system successfully blocks attacker IP addresses, indicating the potential for practical implementation to enhance cybersecurity within the industrial sector.

Kata Kunci : Industri, Keamanan Siber, Intrusion Detection System, Machine Learning, Random Forest, Suricata