کنفرانس بین المللی فناوری اطلاعات و دانش

صفحه اصلی / یازدهمین کنفرانس بین المللی فناوری اطلاعات و دانش

Detection and Identification of Cyber-Attacks in Cyber-Physical Systems Based on Machine Learning Methods

نویسندگان :

Zohre Nasiri Zarandi¹

1- دانشگاه بین‌المللی امام خمینی (ره)

کلمات کلیدی :

کنترل تاب‌آور ، سیستم های سایبر-فیزیکی، حملات سایبری فریب، شبکه عصبی

چکیده :

Cyber-physical systems(cps) have made significant progress in many dynamic applications. However, cyber-attacks are a major threat to these systems.Some of attacks which called deception attacks, inject incorrect information from sensors or controllers. It should be noted that the data generated in these systems is produced in very large number,with so much variety, and high speed, so it is important to use machine learning algorithms to facilitate the analysis and evaluation of data and to identify hidden patterns. In this research,the CPS is modeled as a network of agents that move in union with each other, and one agent is considered as a leader, and the other agents are commanded by the leader. The proposed method in this study is to use the structure of deep neural networks for the detection phase, which should inform the system of the existence of the attack in the initial moments of the attack. The use of resilient control algorithms in the network to isolate the misbehave agent in the leader-follower mechanism has been investigated. In the presented control method, after the attack detection phase with the use of a deep neural network, the control system uses the reputation algorithm to isolate the misbehave agent.