Attacks detection and decision framework based on generative adversarial network approach: Case of vehicular edge computing network,Transactions on Emerging Telecommunications Technologies

当前位置： X-MOL 学术 › Trans. Emerg. Telecommun. Technol. › 论文详情

Our official English website, www.x-mol.net, welcomes your feedback! (Note: you will need to create a separate account there.)

Attacks detection and decision framework based on generative adversarial network approach: Case of vehicular edge computing network
Transactions on Emerging Telecommunications Technologies ( IF 2.5 ) Pub Date : 2020-07-31 , DOI: 10.1002/ett.4073
Hichem Sedjelmaci ₁

Affiliation

The application of artificial intelligence in cyber security context has attracted enormous attention, specifically from industrial sector and this is due to the strength of machine learning algorithms to detect the unknown misbehaviors executed by cyber-attacks. The generative adversarial network (GAN) based on a generator and discriminator systems could be used as defense and attack frameworks to protect and attack respectively its target. In this research article, we propose new attacks detection and decision framework based on GAN algorithm to detect accurately the smart and dangerous attacks. The proposed security framework relies on the collaboration between the generator and discriminator systems to determine the relevant attacks' features and hence detect the cyber-attacks with a high accuracy. We present a case study of the proposed detection and decision framework in a context of a vehicular edge computing network and highlight the experimental results by analyzing specifically, the accuracy defense rate and network latency. In the experimentation results, a maximum number of network attacks are detected with a high accuracy, while low network latency is generated. This result is achieved specifically when the number of iterations is high, that is, the accuracy defense rate is close to 90% and the network latency is close to 1000 milliseconds.

中文翻译：

基于生成对抗网络方法的攻击检测与决策框架：以车载边缘计算网络为例

人工智能在网络安全环境中的应用引起了极大的关注，特别是来自工业部门，这是由于机器学习算法能够检测网络攻击所执行的未知不当行为。基于生成器和判别器系统的生成对抗网络（GAN）可用作防御和攻击框架，分别保护和攻击其目标。在这篇研究文章中，我们提出了基于 GAN 算法的新攻击检测和决策框架，以准确检测智能攻击和危险攻击。所提出的安全框架依赖于生成器和鉴别器系统之间的协作来确定相关攻击的特征，从而以高精度检测网络攻击。我们在车辆边缘计算网络的背景下对所提出的检测和决策框架进行了案例研究，并通过具体分析准确率防御率和网络延迟来突出实验结果。在实验结果中，以高精度检测到最大数量的网络攻击，同时产生低网络延迟。这个结果是专门在迭代次数比较多的时候，也就是准确防御率接近90%，网络时延接近1000毫秒的情况下实现的。同时产生低网络延迟。这个结果是专门在迭代次数比较多的时候，也就是准确防御率接近90%，网络时延接近1000毫秒的情况下实现的。同时产生低网络延迟。这个结果是专门在迭代次数比较多的时候，也就是准确防御率接近90%，网络时延接近1000毫秒的情况下实现的。

更新日期：2020-07-31

点击分享查看原文

点击收藏

阅读更多本刊最新论文本刊介绍/投稿指南11