Thesis (M.Sc.) - Cairo University - Faculty of Computers and Information - Department of Information Technology

Cloud computing has a great usage recently while it provides shared resources between users whether in business field or at level of normal users. It also offers many services and applications to users. The main reason behind widespread of cloud computing is that any user can use cloud computing services from anywhere at any time by using any device with only one condition which is providing internet connection. In cloud computing, there are many algorithms used in scheduling process to assign tasks into appropriating resources and also to meet users' requirements which are representing by taking into consideration QoS in scheduling process. In this thesis, there are many different types of scheduling algorithms will be stated and also will be focused on batch-heuristic type of scheduling algorithms. In this thesis, grouped tasks scheduling algorithm is the proposed algorithm used to assign tasks from different users to approbate services with applying QoS of users to their tasks. The main objective of this algorithm is to achieve minimum execution time to all tasks with low latency to urgent tasks. This algorithm will be compared with both Min-Min algorithm and TS-QoS algorithm. Three simulation programs were built using java to evaluate, test and compare the performance of proposed grouped tasks scheduling algorithm with both Min-Min algorithm and TS-QoS algorithm. There are six performance metrics used in comparison execution time, load balancing, latency of long tasks, latency of tasks with urgent users, latency of tasks with expected urgent priority and latency of expected urgent priority of tasks with urgent users

Issued also as CD