Cloud Computing Resource Scheduling based on Improved Semantic Search Engine

Cloud computing resource has the features of dynamic, heterogeneous, distributed and complexity etc. Meanwhile the numbers of resources and tasks to be scheduled in Cloud are usually variable. This makes the Cloud resource scheduling a complex optimization problem. None of existing Cloud systems is both being an automated scheduling and considering the optimal usage of resources. To address above problems, we propose a Cloud computing resource scheduling strategy using improved Semantic Search Engine (ISSE). SSE is a new type of service search engine developed by Semantic Computing laboratory in University of California, Irvine. It provides Cloud users with a friendly problem-driven interface to automatically schedule resources that would be used to build a solution according to users' requirementsin the aid of semantic information from resources and user requirements. Further we adopt improvedgenetic algorithm (IGA) in SSE to optimize the scheduling so as to obtain the optimal usage of resources. In our proposed IGA there should be a code distant between the selected parents to retain the population diversity and obtain the valid solution. The architecture of our proposed ISSE is presented as well as the process and implementation of IGA. The experiment results show our proposed ISSE is feasible and can reduce about 16% average tasks execution time comparing to the traditional Cloud resource scheduling (TCRS).