An Efficient Co-Resident Dos Attack Defense Mechanism for Cloud Computing using Two-Player Security Game Approach
Rethishkumar S.1, R. Vijayakumar2
1Rethishkumar S, Research Scholar, Department of Computer Sciences, Mahatma Gandhi University, Kottayam (Kerala), India.
2Dr. R. Vijayakumar, Professor, Department of Computer Sciences, Mahatma Gandhi University, Kottayam (Kerala), India.
Manuscript received on December 02, 2017. | Revised Manuscript received on December 29, 2017. | Manuscript published on December 20, 2017.| PP: 8-15 | Volume-4 Issue-7, December 2017 | Retrieval Number: G0779114717/2017©BEIESP
Open Access | Ethics and Policies | Cite
© The Authors. Published By: Blue Eyes Intelligence Engineering and Sciences Publication (BEIESP). This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/)
Abstract: For cloud computing systems, Virtual Machines (VM) were conceived as the basic component. However, VMs give effective computing resources; they were prone to lots of security threats. Whereas few threats can be easily rectified, but few attacks like co-resident attacks were tedious process to identify. So, to reduce the co-resistance DOS attacks by creating it as tedious for attackers to initiate attacks, two-player game approach based defense mechanism is suggested in our work. The attacker behavior variations among the attacker and normal users under PSSF VM allocation policy, is examined initially in the proposed mechanism. EDBSCAN (Enhanced Density-based Spatial Clustering of Applications with Noise), is utilized to do the clustering analysis process. Based on the clustering algorithm, the Partial labeling is performed, to partially comprehend the users as legal or malicious. In order to classify the nodes, the semi-supervised learning using Deterministic Annealing Semi-supervised SVM (DAS3VM) optimized by branch and bounds method is performed. The two-player security game approach helps to raise the cost of introduction new VMs therefore reducing the probability of initiating co-resident DOS attack, once after the user accounts were classified. Therefore, the security threats can be avoided effectively with the help of the proposed defense mechanism. Experimental result confirms that the suggested co-resident DOS attack defense mechanism makes a desirable involvement to avoid the security threats.
Keywords: Co-Resident DOS Attack, PSSF, EDBSCAN, DAS3VM, Branch and Bound Method.