Analisis Performansi Ids Menggunakan Metode Deteksi Anomaly-based Terhadap Serangan Dos

Abstract

Abstrak Intrusion Detection System (IDS) merupakan sistem yang dapat mendeteksi aktivitas yang mencurigakan dalam sebuah jaringan. Metode deteksi Anomaly-Based dipilih agar dapat mendeteksi aktivitas yang mencurigakan dan tidak normal bagi sistem yang tidak dapat dilakukan oleh metode Signatured-based. Pada penelitian ini dilakukan pengujian serangan menggunakan tiga tools DoS yaitu tools LOIC, Torshammer dan Xerxes dengan scenario pengujian yaitu menggunakan IDS serta tanpa IDS. Dari hasil pengujian yang telah dilakukan IDS berhasil mendeteksi serangan yang dikirim, untuk pengiriman paket serangan terbanyak berurutan yaitu Torshammer, Xerxes dan LOIC. Pada pendeteksian tools serangan Torshammer kepada target FTP Server didapatkan sebanyak 9421 paket, untuk tools Xerxes yaitu sebanyak 10618 paket dan tools LOIC sebanyak 6115 paket. Sedangkan serangan kepada target Web Server untuk tools torhsammer sebanyak 299 paket, untuk tools Xerxes sebanyak 530 paket dan untuk tools LOIC sebanyak 103 paket. Akurasi dari hasil performansi IDS yaitu sebesar 88,66%, presisi sebesar 88,58% serta false positive rate sebesar 63,17%. Kata kunci : Intrusion Detection System, Anomaly-Based, Keamanan Jaringan, Denial of Service Abstract Intrusion Detection System (IDS) is a system that can detect suspicious activity in a network. AnomalyBased detection method is chosen to be able to detect suspicious and abnormal activities for the system that cannot be done by Signatured-based methods. In this study, attack testing was carried out using three DoS tools, namely LOIC, Torshammer and Xerxes tools with the testing scenario of using IDS and without IDS. From the results of testing that has been done, IDS has successfully detected the attack sent, for sending the most consecutive attack packages, namely Torshammer, Xerxes and LOIC. In the detection of tools, Torshammer's attack on the FTP Server target was 9421 packages, for Xerxes tools, there were 10618 packages and LOIC tools as many as 6115 packages. While attacks on the target Web Server for torhsammer tools as many as 299 packages, for Xerxes tools as many as 530 packages and for LOIC tools as many as 103 packages. The accuracy of the IDS performance results is 88.66%, precision is 88.58% and the false positive rate is 63.17%. Keywords: Intrusion Detection System, Anomaly-Based, Network Security, Denial of Service