Trusted Cloud Service Selection Model Of Multi-QoS And Applications

As service-oriented computing mode, cloud computing integrates computing resources, storage resources and network resources organically using virtualization technology, and provides customers with the model of multi-tenancy, data outsourcing and pay-on-demand. This kind of service model provides convenience to customers. Meanwhile, it also makes them face many challenges. For example, how to measure, evaluate comprehensively and sort the cloud service s QoS attribute, which could help customer to select the most satisfied cloud services from many cloud services with the same or similar service but different QoS. For security-aware customers, specifically, it is essential to measure the subjective security of cloud services effectively, and select trusted cloud services.The cloud service selection is used as the research object, and its security problem is deemed as the breakthrough point. Based on operations research, fuzzy set theory, probability theory and system decision theory, this dissertation proposes the subjective trust evaluation model based on reputation and risk firstly, which is used to build trusted cloud computing environment. Then, the cloud service subjective security measure model based on subjective trust is designed to measure the subjective security of cloud services effectively. Third, the index system of the cloud service selection based on AHP is designed, and multi-dimension QoS trusted cloud services selection algorithm based on two-step fuzzy comprehensive evaluation is presented, which is used to evaluate and sort different cloud services, thus providing the basis for selecting trusted cloud services to customers. At last, the trusted cloud services selection strategy is applied to the cloud workflow scheduling, and trusted cloud workflow scheduling algorithm is put forward, which is used to select trusted computing services and storage services. Four researching results of this dissertation are listed as follows.(1) The subjective trust evaluation model based on reputation and risk is presentedAiming at the problem of trust crisis caused by multi-tenancy in cloud computing environment, a subjective trust evaluation model based on reputation and risk is put forward, to provide the basis for building trust relationship between cloud service providers and tenants. Considering subjective trust s characteristics of fuzziness, randomness, time-attenuation and risk, the cloud model theory is used to quantify the reputation and risk of the subjective trust, and they are synthesized to synthetic cloud. The value of the synthetic cloud is denoted as the trust of the corresponding entity. Experimental results show that the model improves in the accuracy and effectiveness of the describtion of subjective trust to some extent.(2) The measure model of the cloud services subjective security based on subjective trust is put forwardFor the problem of quantifying subjective security difficultly, the measure model and algorithm of trust subjective security based on the subjective trust is proposed. Different attributes of subjective trust are quantified using the cloud model theory, and the method of determining weight coefficient based on the cloud model and AHP is designed, which is used to calculate the weight coefficient of different attributes. On this basis, different attributes are synthesized into comprehensive cloud through synthetic operator. The value of comprehensive cloud is denoted the subjective security of cloud services. Simulation experiments show that the algorithm has the advantages of fine dispersion and high sensitivity in describing the subjective security of cloud services.(3) The multi-dimension QoS trusted cloud services selection algorithm based on two-step fuzzy comprehensive evaluation is givenTo provide simple and effective decision-making basis for customers, the index system of the cloud service selection based on AHP is designed, and multi-dimension QoS trusted cloud services selection algorithm based on two-step fuzzy comprehensive evaluation is presented. The assessment criteria of cloud services are divided into quantitative and qualitative criteria, different methods are adopted to quantify them respectively, and the secondary fuzzy comprehensive evaluation and ordering of different cloud services are completed by fuzzy synthesis operator. Simulation results show that the algorithm has good performance in normalized discounted cumulative gain and load balancing deviation.(4) Trusted cloud workflow scheduling model and algorithm are put forwardAiming at the security risk in cloud workflow scheduling, trusted cloud workflow scheduling model and algorithm are put forward. First of all, the cloud service subjective security measure model based on subjective trust is used to measure the subjective security of computing services and storage services; Then, taking time, cost, and subjective security into account, cloud workflow scheduling model and the corresponding algorithm based on discrete particle swarm optimization and set covering problem are put forward. They are used to select computing services and storage services, which can satisfy subjective security requirements. Compared with the traditional algorithm, simulation experiments show that this algorithm has improved performance in the time, cost and subjective security.