Network Anomaly Detection Method Based on Immune Bionic Mechanism and Graph Neural Network

doi:10.3969/j.issn.1671-1122.2021.08.002

Abstract

Abstract:

This paper proposes a network anomaly detection method based on immune bionic mechanism and graph neural network by imitating the risk prevention mechanism of biological system, which uses graph neural network to deeply mine the sub graph information near the node. While considering the content features of the network, the structural features based on graph were integrated into the model, which can be used as the basis of anomaly detection in the network, so as to better mine the anomaly information in the network. At the same time, graph representation learning technology was integrated into network anomaly detection to solve the problem of feature representation. Based on CICIDS2017 dataset, Cora dataset and Reddit dataset, the experimental results show that this method can better mine network anomalies and improve the accuracy of anomaly detection.

Key words: network anomaly detection, immune bionic, node, graph neural network, graph representation learning

CLC Number:

TP309

QIN Zhongyuan, HU Ning, FANG Lanting. Network Anomaly Detection Method Based on Immune Bionic Mechanism and Graph Neural Network[J]. Netinfo Security, 2021, 21(8): 10-16.

Figures/Tables 5

References 16

[1]	XIN Zhuang. Research on Intrusion Detection Method Based on Artificial Immune[D]. Guizhou: Guizhou University, 2019.
	辛壮. 基于人工免疫的入侵检测方法的研究[D]. 贵州:贵州大学, 2019.
[2]	ESWARAN D, FALOUTSOS C. Sedanspot: Detecting Anomalies in Edge Streams[EB/OL]. https://ieeexplore.ieee.org/document/8594926 , 2021-01-02.
[3]	ESWARAN D, FALOUTSOS C, GUHA S, et al. Spotlight: Detecting Anomalies in Streaming Graphs[EB/OL]. https://dl.acm.org/doi/10.1145/3219819.3220040 , 2020-12-12.
[4]	PEROZZI B, AL-RFOU R, SKIENA S. Deepwalk: Online Learning of Social Representations[EB/OL]. https://dl.acm.org/doi/10.1145/2623330.2623732 , 2020-11-12.
[5]	WU Zonghan, PAN Shirui, CHEN Fengwen, et al. A Comprehensive Survey on Graph Neural Networks[J]. IEEE Transactions on Neural Networks and Learning Systems, 2021, 32(1):4-24. doi: 10.1109/TNNLS.5962385 URL
[6]	LIU Weiwen, ZHANG Yin, WANG Jianling, et al. Item Relationship Graph Neural Networks for E-commerce[EB/OL]. https://ieeexplore.ieee.org/document/9372806 , 2020-11-12.
[7]	ZHANG Lu, ZHANG Jian, LI Zhibin, et al. Towards Better Graph Representation: Two-branch Collaborative Graph Neural Networks for Multimodal Marketing Intention Detection[EB/OL]. https://ieeexplore.ieee.org/document/9102918 , 2020-11-12.
[8]	WU Yao, SONG Yu, HUANG Hong, et al. Enhancing Graph Neural Networks via Auxiliary Training for Semi-supervised Node Classification[EB/OL]. https://www.sciencedirect.com/science/article/abs/pii/S0950705121001477 , 2021-01-11.
[9]	HENAFF M, BRUNA J, LECUN Y. Deep Convolutional Networks on Graph-structured Data[EB/OL]. https://arxiv.org/abs/1506.05163 , 2020-12-11.
[10]	VELIČKOVIĆ P, CUCURULL G, CASANOVA A, et al. Graph Attention Networks[EB/OL]. https://arxiv.org/abs/1710.10903 , 2020-11-11.
[11]	GILMER J, SCHOENHOLZ S S, RILEY P F, et al. Neural Message Passing for Quantum Chemistry[EB/OL]. https://dl.acm.org/doi/10.5555/3305381.3305512 , 2020-12-11.
[12]	HAMILTON W L, YING R, LESKOVEC J. Inductive Representation Learning on Large Graphs[EB/OL]. https://dl.acm.org/doi/abs/10.5555/3294771.3294869 , 2020-12-11.
[13]	BELGHAZI M I, BARATIN A, RAJESWAR S, et al. Mine: Mutual Information Neural Estimation[EB/OL]. https://arxiv.org/abs/1801.04062 , 2020-11-11.
[14]	HJELM R D, FEDOROV A, LAVOIE-MARCHILDON S, et al. Learning Deep Representa-tions by Mutual Information Estimation and Maximization[EB/OL]. https://arxiv.org/abs/1808.06670 , 2020-11-11.
[15]	GILMER J, SCHOENHOLZ S S, RILEY P F, et al. Neural Message Passing for Quantum Chemistry[EB/OL]. https://dl.acm.org/doi/10.5555/3305381.3305512 , 2020-11-11.
[16]	HAMILTON W L, YING R, LESKOVEC J. Inductive Representation Learning on Large Graphs[EB/OL]. https://dl.acm.org/doi/abs/10.5555/3294771.3294869 , 2020-11-12.

数据集	节点数	边数	特征	类
IDS2017	9015	692703	80	2
Cora	2708	5429	1433	7
Reddit	231443	11606919	602	41

方法	IDS2017	Cora	Reddit
DeepWalk	71.2%	70.3%	32.4%
GCN	92.4%	78.8%	93.7%
GraphSage	73.1%	75.1%	89.2%
本文方法	98.7%	81.7%	94.1%