Authors: Farzad Amirjavid, Hamid Mcheick, Mohamed Dbouk
Addresses: Department of Computer Science and Mathematics, University of Quebec at Chicoutimi (UQAC), 555 Boul de l'Universite, Quebec, G7H 2B1, Canada. ' Department of Computer Science and Mathematics, University of Quebec at Chicoutimi (UQAC), 555 Boul de l'Universite, Quebec, G7H 2B1, Canada. ' Faculty of Science (I), Lebanese University, Hadath Campus, Beirut, Lebanon
Abstract: Requests for services and user demands increase continuously and faster than technology development. To keep the service offering time reasonable, we can decrease the servers total elapsed time (STET) instead of waiting for newer technologies. This paper focus on division of jobs between computers operating in network to improve the computing speed. Furthermore, we deal with the problem concerning how to arrange services in a way that the STET will be minimised in the best way. Service oriented architecture principals are used to participate jobs between computers. We have focused on time parameter to design software at the software architecture level; we follow the goal to minimise the STET. A mathematical model is proposed to express computing time relations and an LP model is applied to find best combination of services regarding computing speed. A case study is presented to demonstrate the time relations.
Keywords: software architecture; computing speed; service oriented architecture; SOA; separation of concerns; job division; total elapsed time; software design; mathematical modelling.
International Journal of Communication Networks and Distributed Systems, 2011 Vol.6 No.1, pp.97 - 113
Published online: 03 Dec 2010 *Full-text access for editors Access for subscribers Purchase this article Comment on this article