動態範圍抗體為基礎之類免疫演算法於網路入侵偵測問題

Dynamic Range Antibody Based Artificial Immune System for Network Intrusion Detection Problems

葉昱圻(Yu-Chi Yeh)

指導教授 ： 張百棧

元智大學/資訊學院/資訊管理學系/碩士(2013年)

https://doi.org/10.6838/YZU.2013.00098

摘要

隨著電腦運算能力的提升與網際網路的迅速發展，網路異常入侵偵測系統扮演著資訊安全中相當重要的角色。本研究希望透過啟發式演算法中的類免疫演算法為基礎，希望透過免疫系統中，記憶與學習的特性發展出一個類免疫分類演算法來解決異常入侵等分類問題，演算法中包含具變動範圍抗體之生成、抑制細胞、記憶細胞等機制，其中記憶細胞是本研究中完成建模後之分類器，其負責找出並產生可消滅抗原之抗體，透過這些設計希望提高整體演算法的效率及穩定性。
本研究實驗設計以網路異常偵測為研究方向，並以KDD Cup 1999資料集做為本研究的主要資料，根據實驗結果顯示本研究所提之類免疫演算法結合具變動範圍抗體生成等機制，可在異常入侵偵測問題上得到相當好的分類效果。相較於其他分類演算法，本研究所建立之類免疫分類演算法具有相當程度的正確性及穩定性。

關鍵字

類免疫演算法 ； KDD’ 99 ； 分類問題 ； 入侵偵測

並列摘要

With the development of computer science and network technology, anomaly intrusion detection systems play a very important role in network security. This study tries to construct a method based on a kind of heuristic algorithms, artificial immune system. According to the memory and learning characteristic of the immune system to develop an novel artificial immune classification algorithm and solve classification problems. The algorithm contains variable antibody production, suppressor cells, memory cells, etc. In which, the memory cell is the classifier which is completed training which is responsible to produce the antibodies to identify the antigens in this study. We hope that through to design these mechanisms to improve the overall efficiency and stability in classification problems.
In experimental result, the instance of network intrusion detection problem we test is KDD Cup 1999 data set, and the algorithm we proposed can capture high accuracy when an intrusion connection attacks better than other AIS based classifier and machine learning.

並列關鍵字

Artificial Immune Systems ； KDD' 99 ； Classification Problems ； Intrusion Detection

參考文獻

30. 陳炫明，「發展漸進式學習之類免疫演算法於分類問題」，元智大學，碩士論文，2012。

2. H. Jabeen, A. R. Baig, “Two Layered Genetic Programming for Mixed-attribute Data Classification”, Applied Soft Computing, 12, 1, pp. 416-422, 2012.

4. R. Bace, “An Introduction to Intrusion Detection and Assessment: For System and Network Security Management”, ICSA White Paper, 1998.

8. S. Ghodratnama, M. R. Moosavi, M. Taheri, M. Zolghadri Jahromi, “A Cost Sensitive Learning Algorithm for Intrusion Detection”, Iranian Conference on Electrical Engineering (ICEE), Isfahan, Iran , pp. 559-565, 2010.

9. P. Sangkatsanee, N. Wattanapongsakorn, C. Charnsripinyo, “Practical Real-time Intrusion Detection using Machine Learning Approaches”, Computer Communications, 34, pp. 2227-2235, 2011.