Abstract
Distributed denial of service (DDOS) attack constitutes one of the prominent cyber threats and among the hardest security problems in modern cyber world. This research work focuses on reviewing DDOS detection techniques and developing a numeric stable theoretical framework used for detecting various DDOS attacks in cloud. Main sections in the paper are devoted to review and analysis of algorithms used for detection of DDOS attacks. The framework theorized here deals with the variability calculation method in conjunction with sampling, searching methods to find a current state of a particular parameter under observation for detecting DDOS attacks. This way a solution is to build that measure the performance and conduct the monitoring framework to capture adversity related to DDOS attacks. The described algorithm intends to capture the current context value of the parameters that determine the reliability of the detection algorithm and the online pass algorithm helps to maintain the variability of those collected values thus maintaining numerical stability by doing robust statistical operations at endpoints of traffic in cloud based network.
Highlights
Internet of Things (IoT) has evolved in modern times in leap and bounds
[14] Jun-Ho Lee, Min-Woo Park, Jung-Ho Eom, and TaiMyoung Chung in their paper ―Multi-level intrusion detection system and log management in cloud computing‖ proposed multi-level Intrusion detection systems (IDS) in combination with log management approach to strengthen the security in cloud based network so that anomaly behavior can be detected in cloud environments
It is hard to identify the intensity of attack, if the attack is slow in nature and has discrete events occurring based on demand cloud services
Summary
Internet of Things (IoT) has evolved in modern times in leap and bounds. Incidents of attacks over the Internet especially DDOS attacks [1] are increasing day by day. The cloud based networks [16] and services are prone to suffer from malicious attacks because of their inherent characteristics of being accessible globally any time and due to the frequent changes in topology and development of IoT as well as because of landscape nature of Internet. It is known fact that both the internal and external anatomy [8] of the data-center matters, how it is structured architecturally to measure the volume of traffic is the main critical point, if somehow the intruders are able to launch a slow attack it must be detectable or if it is a sudden flood of packets the system must be able to mitigate the flood to have clean traffic This is not possible unless there is continuous monitoring which includes the mapping of threats [18] cope with the understanding correlations of all the factors contributing to the adversity.
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
