International Journal of Computer Applications (0975 – 8887) Volume 15– No.2, February 2011 32 An Efficient Economic Model User-Oriented for Cloud Computing Ghalem Belalem Dept. of Computer Science Faculty of Sciences University of Oran - Algeria BP. 1524, EL M’Naouer, Oran, Algeria (31000) Samah Bouamama Dept. of Computer Science Faculty of Sciences University of Oran - Algeria BP. 1524, EL M’Naouer, Oran, Algeria (31000) Larbi Sekhri Dept. of Computer Science Faculty of Sciences University of Oran - Algeria BP. 1524, EL M’Naouer, Oran, Algeria (31000) ABSTRACT The Cloud computing offers the availability and performance of services those users require to resolve their requests in reduced time and lower cost. Several studies have overcome this problem by the proposed algorithms borrowed from economic models of real world economy to ensure that quality of service. The work presented in this paper allows in the first time to satisfy customers in terms of treatment quality of cloudlets cheaply in the cloud, while offering improved algorithms auction of simulator GridSim specific of CloudSim to simulate economic environment in the cloud, in order to provide virtualized resources. In the second time, we propose a hybrid business model that behaves well relative to the different models studied. General Terms Cloud computing, Grid computing, Distributed systems, Resource management. Keywords Cloud computing, CloudSim, Auction model, Cloudlet, Datacenter. 1. INTRODUCTION Today, applications are becoming more and more resource- intensive, either in memory or computing power. These resources are limited on a machine, the need to distribute applications over a network of computers, local or remote, is obvious. Grid computing is mainly focused on the provision of a significant amount of shared resources for intensive computing (research or physical applications). The protocols used are tailored to the specific topology of the network (low latency, high throughput). Cloud Computing is based instead on the provision of many services and data to users [1], without their having to manage the complex infrastructure required. Because of its support web services, it uses Internet protocols and is therefore subject to the constraints of the latter. The cloud computing environment suggests a future where we do not calculate on local computers, but on equipment centralized computing and storage operated by third parties. This environment has great flexibility and ease of use, availability of data and services became one of the biggest problems to address and improve. With cloud computing, companies can raise large capacity in an instant without having to invest in new infrastructure, new staff or new software. This model of development was born after the giants of the Internet have created their own infrastructure to meet their own needs; they are now able to ensure their development, but also to share resources at competitive prices. Cloud infrastructure modeling and simulation toolkits should provide support to economic entities such as brokers and exchanges Cloud for the negotiation of services between customers and suppliers. Among the currently available simulators who have an interest in our work, the simulator GridSim which provides management support and economic resources CloudSim simulator [8] [10] to model the cloud. Our job is to propose a hybridization of simulators CloudSim and GridSim aspects, which are implementing economic models of GridSim and virtualization of computing resources of CloudSim and then consider four auction models and their improvement. Finally, propose an economic model that brings together all beneficial criteria on the client side of the auction models. The rest of the paper is structured as follows: in Section 2, we define cloud computing environments as an important scientific trend. Section 3 is dedicated to the simulator GridSim; define the simulator and its uses in the Grid. We present in Section 4, the simulator CloudSim, The following section identifies changes to CloudSim to ensure proper management of resources in the simulator CloudSim. The different experiments are presented in Section 6. We end our paper with a summary and some extension work that we will consider doing so. 2. CLOUD COMPUTING Cloud computing can be defined as "A type of parallel and distributed system consisting of a collection of interconnected computers and they are virtualized and dynamically generated and submitted as one or more computing resources based on service level agreement established through negotiation between the provider and consumers" [9]. Some examples of new infrastructure are Microsoft Azure Cloud Computing [5], Amazon EC2, Google and Aneka [6]. Computing power in cloud computing environments is provided by a collection of data centers, which are typically installed with hundreds of thousands of servers [3]. The layered architecture of a typical cloud-based data centers is shown in “Fig. 1,” In the lower layers there are huge hardware resources (storage and