A Novel Framework for DDoS Detectionin Huge Scale Networks, Thanksto QoS Features

It is not been a long time since the advent of cloud-based technology. However, in this short period of time several advantages and disadvantages have been emerged. This is a problem solving technology with some threats as well. These threats and potential damages are not only limited to the cloud-based technologies, but they have always been against computer network infrastructures. One of these examples is Distributed Denial-of-Service (DDoS) intrusion which is of course one of the most complex and the most dangerous types of attacks. The impact of this type of attack, due to its powerful nature, is much higher on cloud systems since in case of occurrence, the service providers lose their services completely as well as their reputation and loyal customers. This, apparently, can even lead to the collapse of the stock and other destructive consequences. On the other hand, due to the properties of cloud service providers including large-scale infrastructures, DDoS intrusion detection algorithms need high sensitivity, innovation, and general improvements. Traditional structures of DDoS attack detection algorithms are designed for small-scale networks or at most for application camps. Lack of efficient algorithm is seemingly apparent for the large-scale networks. Therefore, in this context we utilize standard methods as well as a proposed hybrid protocol which is more appropriate in connection with cloud structures in order to detect DDoS attacks. Keywords-Cloud Computing, DDoS Attacks, Intrusion Detection Algorithms, Large-scale networks I. LITERATURE REVIEW Many studies have been performed on cloud computing service providers. One of the most important sections of these studies is to review different challenges in these structures. For instance, a research has been conducted by the IDC in 2009, and after a full investigation among users, they have come to the conclusion that the most important challenge from the perspective of users of cloud services is security [1]. On the other hand, regarding the research performed by Lockheed Martin [2], a major challenge after data security is Intrusion Detection in the cloud structures. So we can guess how critical and important communication security is even in this special case. On the other hand, several studies have performed on the centralized threats of cloud structures. Cloud services and other network technologies may have many threats as well [3]. For example, we can mention the forgery attacks, unauthorized access attacks, and DDoS. In this research we focus on DDoS attacks against cloud computing service providers. In these attacks, the attacker tries to fully utilize the resources required by users in order to prevent their access. These types of attacks can be occurred either in a distributed or a specific resource [5]. For example, in a cyber-attack occurred in 2011 against the Amazon EC2 service, Sony Online Entertainment segment was impaired and unavailable [5] which also caused in enormity losses. With the emergence of cloud technology, researchers’ attention in availability is increased much more rather than in information security [6] with regard to the fact that the basis of this technology is the high availability. Thus, several algorithms and methods are proposed to prevent attacks against availability [3, 4, 6, 7, 8, 9, 10, 11, and 12] each has proposed his own algorithm structure. For instance, [13] and [14] put ways of detecting an attack together and compared themgenerally. In addition, [15] evaluated the quality of the methods in intrusion detection alerts. Apparently, many techniques have been proposed for intrusion detection and they have all been studied by other researchers.