UBP-Trust: User Behavioral Pattern Based Secure Trust Model for Mitigating Denial of Service Attacks in Software as a Service (SaaS) Cloud Environment

Abstract

The problem of security enforcement in cloud environment has been discussed in number of situations and the most approaches uses minimum number of features to mitigate the denial of service attacks in cloud environment. The methods suffers with the problem of poor detection accuracy and false classification ratio, to overcome the issue, we propose a novel approach to mitigate the denial of service attacks in SaaS layer of cloud environment. This paper discusses a UBP-Trust model, which monitors the behavioral patterns of the users of cloud environment at different situations. Based on the monitored results, the method generates user behavior pattern which represents, the number of times the user has accessed the service, the number of times the service has been accessed and finished successfully, the amount of data being sent, the number of false invocation, the variance of protocol and so on. Using all these features considered the method generates the behavioral pattern and used to compute the user trust weight for each user being monitored. Based on the weight computed, he will be decided as malicious or genuine and based on which the method restrict the user from accessing the service. The proposed method produces efficient results in DDOS detection accuracy and produces less time complexity and false classification ratio.