ارسل ملاحظاتك

ارسل ملاحظاتك لنا







An Intrusion Detection System for DOS Attacks Based on Neural Networks

المصدر: مجلة العلوم الاقتصادية والسياسية
الناشر: الجامعة الأسمرية الإسلامية - كلية الاقتصاد والتجارة
المؤلف الرئيسي: Bentaher, Omran Ali (Author)
مؤلفين آخرين: Al Bhbah, Atia M. (Co-Author)
المجلد/العدد: ع7
محكمة: نعم
الدولة: ليبيا
التاريخ الميلادي: 2016
الشهر: يونيو
الصفحات: 430 - 445
رقم MD: 765769
نوع المحتوى: بحوث ومقالات
اللغة: الإنجليزية
قواعد المعلومات: EcoLink
مواضيع:
كلمات المؤلف المفتاحية:
Denial-of-Service Attack | Feature Selection | Intrusion Detection Systems | Neural Networks
رابط المحتوى:
صورة الغلاف QR قانون

عدد مرات التحميل

12

حفظ في:
المستخلص: Intrusion detection systems (IDSs) have become an essential component of computer security to detect attacks that occur despite the best preventive measures. A problem with majority of current intrusion detection systems is their rule-based nature. In this paper, we propose an optimized neural network based IDS for detecting DoS attacks. The proposed system consists of Multiple Layered Perceptron (MLP) decision block and a feature reduction preprocessing subsystem. The system was optimized and tested on benchmark KDDCUP’ 99 dataset. Several experiments have been conducted to choose the important features from full set of 41, based on three factors: training time, testing time and detection accuracy. Final optimized MLP IDS provides superior accuracy of 98.5%, substantially better than other referential IDS systems published up to now.

عناصر مشابهة