Main Article Content
The objective of the paper is to discuss the proposed dynamic software model to detect and prevent intrusion in the cloud network.
The Behavior Profiling Algorithm (BPA) has been used to detect the intrusion in cloud network. For finding the intruder in the network the Event Log Entries and the network Unique Identification Address (UIA) has been fetched from the server and then the collected attribute values have been transferred to prevention module. In the prevention module the dynamic statistical approach model has been used to prevent the network systems and data which are available in the Cloud Network.
For testing the proposed model the 100 cloud network systems were taken and based on the loss of packets (in MB) ranges the samples were classified as 0-100, 101-200, 201-300, 301-400, 401-500, 501-600, 601-700 respectively. The range of data loss is assumed to be an interval of 100 Mbps. It is assumed that the higher the data loss ranges, the more data is lost. The mean, variance, and standard deviation were calculated to verify the data loss ranges. The mean (average) of the data loss in the ranges 0-100 is 060.77 and the mean in the ranges 101-200 is 144.714 data losses, which gradually increases in proportion to the data loss ranges, and in the ranges 601-700 it is 665.769 data losses. From the statistical approach model, the differences between mean and variance indicated that the intruder attacked the files during the data transformation in the network. Therefore, the administrator has to monitor the warning message from the proposed IPS model and get data packet losses in the transformation. If the frequency of data loss is low, the administrator can assume that the data flow is low due to network problems. On the other hand, if the frequency of data loss in the network system is high, he can block the transformation and protect the data file. This paper concludes that the behavioral profiling algorithm combined with a statistical model achieves an efficiency of over 96% in wired networks, over 97.6% in wireless networks, and over 98.7% in cloud networks.
In the previous paper discussed the approach which has been implemented with 40 nodes and the result of the proposed algorithm produced above 90%, 96% and 98% in the wired, wireless and cloud network respectively. Now, the model has been implemented with 100 nodes the result has been increased. This study concluded that, the efficient algorithm to detect the intrusion is behaviour profiling algorithm, while join with the statistical approach model, it produces efficient result.