Ardi Rahman Sidiq, - (2024) PERANCANGAN MODEL DETEKSI DDOS MENGGUNAKAN ALGORITMA RANDOM FOREST DENGAN PEMBERITAHUAN CEPAT MELALUI TELEGRAM. S1 thesis, Universitas Pendidikan Indonesia.
![[img]](http://repository.upi.edu/style/images/fileicons/text.png) |
Text
S_TEKOM_2009780_Title.pdf Download (760kB) |
|
Text
S_TEKOM_2009780_Chapter 1.pdf Download (268kB) |
|
Text
S_TEKOM_2009780_Chapter 2.pdf Restricted to Staf Perpustakaan Download (462kB) |
|
Text
S_TEKOM_2009780_Chapter 3.pdf Download (508kB) |
|
Text
S_TEKOM_2009780_Chapter 4.pdf Restricted to Staf Perpustakaan Download (717kB) |
|
Text
S_TEKOM_2009780_Chapter 5.pdf Download (256kB) |
|
Text
S_TEKOM_2009780_Appendix.pdf Restricted to Staf Perpustakaan Download (786kB) |
Abstract
Melindungi sistem dari serangan cyber dan upaya tidak sah, termasuk penyusupan atau pemindaian, adalah fokus utama. Di Indonesia, hal ini diatur oleh Undang-Undang Nomor 11 Tahun 2008 tentang Informasi dan Transaksi Elektronik (UU ITE). Serangan DDoS adalah ancaman serius bagi keamanan jaringan, di mana penyerang mencoba menonaktifkan layanan dengan membanjiri target dengan lalu lintas besar. Penelitian ini bertujuan mengembangkan model deteksi DDoS menggunakan algoritma Random Forest dan dataset CICDDoS2019, mencakup 431.371 sampel, dengan 345.096 sampel (80%) untuk pelatihan dan 86.275 sampel (20%) untuk pengujian. Random Forest dibandingkan dengan SVM, KNN, dan LSTM untuk memastikan metode yang tepat. Berdasarkan pengukuran kinerja keempat metode tersebut, Random Forest dipilih karena akurasi 93%, meskipun sama tinggi dengan SVM, tetapi unggul dalam generalisasi dan mengurangi overfitting melalui banyak pohon keputusan, dibandingkan dengan KNN (92%) dan LSTM (82%). Proses penelitian meliputi pengumpulan data, pra-pemrosesan, dan ekstraksi fitur seperti volume lalu lintas (Gbps), jumlah paket per detik (pps), dan permintaan per detik (rps). Generated data juga digunakan, dengan variasi dalam 72 parameter serangan, termasuk Protocol (nilai 6 untuk TCP dan 17 untuk UDP), Flow Duration (2.000.000 hingga 5.000.000 mikrodetik), Total Fwd Packets (5.000 hingga 20.000 paket), dan Flow Bytes/s (8.000.000 hingga 20.000.000 byte/s) serta mengintegrasikan notifikasi cepat melalui Telegram. Evaluasi model menunjukkan akurasi, presisi, recall, dan f1 score sebesar 93% dalam mengklasifikasikan serangan DDoS dan rata-rata waktu pengiriman notifikasi 0,0174 detik. Penelitian ini berhasil merancang model Random Forest dan mampu mengidentifikasi serangan DDoS, serta dapat diintegrasikan dengan Telegram menggunakan bahasa pemrograman Python untuk notifikasi. ------------- Protecting systems from cyber attacks and unauthorized attempts, including intrusion or scanning, is a primary focus. In Indonesia, this is regulated by Law Number 11 of 2008 concerning Information and Electronic Transactions (UU ITE). DDoS attacks are a serious threat to network security, where attackers try to disable services by flooding targets with large traffic. This research aims to develop a DDoS detection model using the Random Forest algorithm and the CICDDoS2019 dataset, covering 431,371 samples, with 345,096 samples (80%) for training and 86,275 samples (20%) for testing. Random Forest is compared with SVM, KNN, and LSTM to ensure the appropriateness of the method. Based on performance measurements of the four methods, Random Forest was chosen because of its 93% accuracy, although it is the same as SVM, but is superior in generalization and reducing overfitting over many decision trees, compared to KNN (92%) and LSTM (82%). The research process includes data collection, pre-processing, and feature extraction such as traffic volume (Gbps), number of packets per second (pps), and requests per second (rps). Generated data is also used, with variations in 72 attack parameters, including Protocol (value 6 for TCP and 17 for UDP), Flow Duration (2,000,000 to 5,000,000 microseconds), Total Fwd Packets (5,000 to 20,000 packets), and Flow Bytes/s (8,000,000 to 20,000,000 bytes/s) as well as integrating fast notifications via Telegram. Model evaluation shows accuracy, precision, recall and f1 score of 93% in classifying DDoS attacks and an average notification sending time of 0.0174 seconds. This research succeeded in designing a Random Forest model and was able to identify DDoS attacks, and can be integrated with Telegram using the Python programming language for notifications.
| Item Type: | Thesis (S1) |
|---|---|
| Additional Information: | https://scholar.google.com/citations?view_op=new_profile&hl=en ID SINTA Dosen Pembimbing: Muhammad Taufik Dwi Putra 6745726 Munawir 6745899 |
| Uncontrolled Keywords: | Random Forest, DDoS, deteksi serangan, Telegram, keamanan jaringan, Random Forest, DDoS, attack detection, Telegram, network security |
| Subjects: | Q Science > QA Mathematics Q Science > QA Mathematics > QA75 Electronic computers. Computer science Q Science > QA Mathematics > QA76 Computer software T Technology > T Technology (General) |
| Divisions: | UPI Kampus cibiru > S1 Teknik Komputer |
| Depositing User: | ARDI RAHMAN SIDIQ |
| Date Deposited: | 23 Sep 2024 08:23 |
| Last Modified: | 23 Sep 2024 08:23 |
| URI: | http://repository.upi.edu/id/eprint/122500 |
Actions (login required)
 |
View Item |