M.A. Thesis:

ob scheduling in the cloud using a meta-heuristic algorithm

Scheduling, resource allocation in cloud computing is of paramount importance; a resource allocation system in cloud computing can be considered as any mechanism that aims to meet the requirements of applications. In addition, the resource allocation mechanism should examine the current state of each resource in the cloud environment in order to provide algorithms for better scheduling and allocation of physical resources or virtual resources and thus reduce operating costs in the cloud environment. Accordingly, in this study, an optimal solution based on cat swarm optimization algorithm for scheduling tasks in the cloud infrastructure is being presented. In this solution, a set of cats will be used, some of which are in search mode and some of which are in tracking mode at the same time. Each cat represents a task-resource mapping and is updated based on the condition of the cat. Assigning a fitting amount to cats minimizes the cost of mapping. Each time a repetition is made, a set of new cats is selected to be in the search mode. Finally, the best location among cats will actually represent the best mapping, which creates the lowest cost among other mappings. Ultimately, in order to review and evaluate the algorithm, the proposed solution was compared with two scheduling solutions based on optimization Gray wolves (GWO) and ant colony (ACO and it was shown that it is more efficient than both solutions, and we have achieved the goal of reducing execution time and energy consumption. So that the amount of energy consumption compared to the ant colony has been reduced by four kilowatt hours and compared to the GWO schedule, there has been a reduction of more than two kilowatt hours. Execution time was also reduced by 30% compared to ant colony-based scheduling and execution time is reduced by 15% compared to GWO. Another parameter that has been considered is the degree of breach of service level agreement. (SLA). In fact, this agreement is the basis for determining the expected level of service that the proposed solution compared to the other two solutions has made the servers more reliable.

Keywords: Keywords: scheduling and resource allocation, cat swarm optimization algorithm, service level agreem