Distributed System for Detection and Prevention Network Attacks to Cloud Computing

The paper describes the problem of detection of intrusion for distributed systems and cloud computing. The goal is to detect and prevent both classical and distributed network attacks such as Denial of Service (DoS, DDoS). The paper identifies a number of problems of various popular cloud computing systems that represent a danger not only obtaining access to the user data, but also it could compromise the integrity and efficiency of the computer system. As solution it is proposed to develop a system for detecting and preventing network attacks. The system consists of several modules designed to perform different functions: detection and prevention of attacks, interaction of system modules, data management and storage. Recognition algorithms are based on the methods of artificial intelligence and the theory of probability. The new solution uses some intellectual methods to recognize attacks as counter to signature-based approach. This paper describes the architecture and functioning of the proposed solutions. It presents the advantages and disadvantages of the described approach. The solution results were presented in the conclusion of the paper. Solution was tested with different types of network attacks on a specially prepared experimental stand.

cloud, security, network attack, artificial intelligence