With the development of mobile communications and the Internet of Things (IoT), IoT devices have increased, allowing their application in numerous areas of Industry 4.0. Applications on IoT devices are time sensitive and require a low response time, making reducing latency in IoT networks an essential task. However, it needs to be emphasized that data production and consumption are interdependent, so when designing the implementation of a fog network, it is crucial to consider criteria other than latency. Defining the strategy to deploy these nodes based on different criteria and sub-criteria is a challenging optimization problem, as the amount of possibilities is immense. This work aims to simulate a hybrid network of sensors related to public transport in the city of São Carlos - SP using Contiki-NG to select the most suitable place to deploy an IoT sensor network. Performance tests were carried out on five analyzed scenarios, and we collected the transmitted data based on criteria corresponding to devices, applications, and network communication on which we applied Multiple Attribute Decision Making (MADM) algorithms to generate a multicriteria decision ranking. The results show that based on the TOPSIS and VIKOR decision-making algorithms, scenario four is the most viable among those analyzed. This approach makes it feasible to optimally select the best option among different possibilities.

Cloud computing is a revolutionary paradigm in relation to its characteristics which can be highlighted the high connectivity, elasticity, assignment of measurable services and ease of management. These attributes favor the promotion of this technology, such as the increasing use of virtualization services in public and private institutions. Therefore, part of the efficiency of this service is based on the allocation of virtual resources that enable different configurations in different applications. In relation to this context, this paper aims to present the allocation module of the Architecture Management of Hybrid Resources in Private cloud Architecture (MoHRiPA), which was modeled with the purpose of providing high performance during the allocation of virtual machines. Its architecture and component specification will be presented, with emphasis on the resource allocation module that has hybrid properties, which makes it possible to support the use of different virtualization tools. The results obtained are described by means of a complete factorial planning with their respective more accurate analyzes about the functionalities and peculiarities of the allocation module.