A Trust Based Security of Resources in Cloud Environment

Abstract

Cloud computing has been attracting the attention of several researchers both in the academia and the industry as it provides many opportunities for organizations by offering a range of computing services. Before cloud computing to become acceptable to everybody both the enterprises and individuals, several issues have to be solved. One of the most important aspects that need to be paid special attention is the cloud security. Trust management is one of the important components in the cloud security that needs special attention. The trust management systems proposed for cloud computing by various researchers have been studied with special emphasis towards their capability and applicability in a practical heterogeneous cloud environment besides implementabilty. An effective trust management system helps cloud service providers and consumers to reap the benefits brought about by cloud computing technologies. Despite the benefits of trust management, several issues related to general trust assessment mechanisms, distrusted feedbacks, poor identification of feedbacks, privacy of participants and the lack of feedbacks integration still need to be addressed. Traditional trust management approaches such as the use of Service Level Agreement are inadequate for complex cloud environments. Due to the multiple vulnerabilities like identification, privacy, personalization, integration, security, and scalability in the existing models, it is proposed for a strong trust model which would create a strong trust between the entities or resources of the cloud. To build a strong trust a strong trust path is necessary between the entities where all the entities in cloud and cloud computing environment would trust each other and the entities that have communication would have valid trust on each other. A mathematical model was proposed to calculate basic trust, dynamic trust and trust for migration.Basic trust was calculated using entropy. Based on the initial trust of the isystem, a new trust for successful transactions was calculated as dynamic trust.The trust models proposed were implemented using Family Gene Genetic Algorithm. The algorithm gives an optimal solution for a large set of data. The implementation of proposed model using this adapted algorithm showed that the resources on a cloud with a strong trust value would always be available for performing any successful transaction. We have proposed a end-to-end trust model which calculates trust based on four parameters namely: utilization, saturation, failure rate and availability. In this model we simulated the results using Monte Carlo method to check with the trust decision making policy. We found from the results that the trust decision is high or low based on the availability of the resources. Based on the trust model and the adapted algorithm the performance of the system using perceived factors were evaluated. The implementation on two different cloud platforms, namely Aneka and Opennebula showed that the model would give better results in terms of Process Time, System Time and Compute Time. Thus we conclude that our model proposes a strong trust path between the entities or resources of the cloud.