Proceedings of 2016 IEEE 13th International Conference on Networking, Sensing, and Control Mexico City, Mexico, April 28-30, 2016 978-1-4673-9975-3/16/$31.00 ©2016 IEEE Predictive Model to Determine Quality of Service on Cloud Computing: Service Dependence Graph Aldo Benhumea Peña, Leticia Dávila-Nicanor, Carlos Omar González-Morán Centro Universitario UAEM Valle de México Atizapán de Zaragoza, Predio San Javier s/n, C.P. 54466. Estado de México, México abenhumeap057@alumno.uaemex.mx, ldavilan@uaemex.mx, cogonzalezm@uaemex.mx Giner Alor-Hernández Division of Research and Postgraduate Studies Instituto Tecnológico de Orizaba Orizaba Veracruz, México galor@itorizaba.edu.mx Abstract— Absence of standards in technology providers such as Cloud Computing and Software as a Service (SaaS), becomes a clear risk factor that may involve their investments, while users may represent a constraint to effective competition within a market, and therefore their ability to choose as service applications and platforms. Therefore it is necessary to establish a risk management program that is flexible enough to deal with an environment variable and evolving risk. One of the main problems experienced by Cloud Computing services is the lack of documentation on the design of infrastructure, and the hardware used for the storage of information, so that the users are not sure how or where they are stored their data for example. This work aims to generate a model of assessment of the availability and efficiency within the available applications in the SaaS in order to establish priority of use through a statistical simulation and graphs. This needs be analyzing the non-functional requirements of the applications available in the SaaS which in turn make indirect use of Infrastructure as a Service (IaaS). Thus our model is projected as a guide that provides us design parameters for its evaluation and the availability based on the evolution of applications from the point of view of the user. To create this model, it is necessary to take into account an analysis of potential risks during application execution and data analysis consultation providing users, in order to efficiently manage resources in an organization. Keywords— Statistical simulation. Assessment SaaS. Quality of service) I. INTRODUCTION One of the main reasons that have slowed the adoption of Cloud Computing and SaaS special use by the public and private industry is the problem of application availability. Architectures of Information Technology (IT) in organizations were focused on hardware, software and network standards, where its limitations relate to the goals set by using software systems. However, the complexity of applications and inconsistent approaches are used to develop, analyze or update such projects that generate incorrect development environments and these are considered the main source of potential risks in project failure. That is the reason why poor architectures reflect low availability, difficulties in maintenance and questionable efficiency of the SaaS applications [1]. The challenge of security in Cloud Computing has been a recurring theme to cover different aspects such as: security shouldn´t only include aspects of access, encryption codes, anonymity, privacy, backups, the recognition of standards and safety that they impose, to the acquisition of responsibilities and dispute resolution, control of services and processes for using Cloud Computing [2]. Strategically, it’s necessary to use international safety standards issued by the Institute of Electrical and Electronics Engineers (IEEE) and the National Institute of Standard and Technology (NIST) in order to project additional security and interoperability, to achieve an impact in the long-term within Cloud Computing as a model reliable. [3]. The use and recruitment of applications in the SaaS represents a potential risk because until today there isn’t a tool that allows us to ensure the performance, efficiency, quality and use of the services that will pay a company. The boundaries of organizational networks are blurred. It generates global environments that allow greater access to its customers, partners and suppliers. The "functionality" of services currently is idealized as a set of components and data associated with each other to achieve faster application response, which is projected at a good decision. However, there are risks, under the proposal to develop new and better applications that may be accessible through the internet and mobile devices capable of using SaaS for improvements in organizational processes, as the IT landscape is full of boundaries or political that are not met, incomplete processes out of implementing recommended practices and standards. This proposal consider the use of Software Reliability Engineering (SRE) that is based on the principles of software development, as this is considered a world economic activity. This approach is used because the main goal of our research is to provide users with tools to assess the quality of service based on the principles of SRE and thus test the feasibility of using SaaS applications to meet needs concrete within organizations, as well as serving as support in decision- making. Because at this stage the operational profiles are defined, operational and probability of occurrence, all these activities are based on the software development process.