Selecting Supernodes in Cloud Fog with a Trust and Reputation Model

/An Exploratory Study of Uncertainty Quantication Applied to the Assessment of an ABM Output

 

Local: Sala Multiuso CIC

Horário: 14h

Palestrante: Thiago Milo (mestrando)

Título: Selecting Supernodes in Cloud Fog with a Trust and Reputation Model

Resumo: The cloud computing model plays a central role providing on-demand processing and storage to online service. This model works very well to almost all digital services. However for some applications low latency communication and sensitive response time are crucial. In this case the centralization of computing and storage capacities promoted by the cloud model could lead to a worst quality of service. To face some of these issues the fog computing model seeks to spread the cloud capacities to the network edge. Some of these fog computing models promotes the participation of end users as local service providers, as fog nodes on the network edge. In this context this paper presents and multiagent approach to the fog computing paradigm. We define a multiagent framework for coordinating devices that works as fog nodes and end devices that consume services from the fog. To this end we proposed the utilization of a trust and reputation model and we show that this proposal could bring good outcomes.

 

Horário: 14h30

Palestrante: Carolina Gonçalves Abreu (doutoranda)

Título: An Exploratory Study of Uncertainty Quantication Applied to the Assessment of an ABM Output

Resumo: Computer models have become an indispensable tool in geophysics and ecology for simulating the behavior of complex dynamical systems. However, computer simulations inherently involve uncertainties. Therefore, identifying and quantifying those uncertainties is critical to enable researchers to make precise statements about the degree of confidence they have in their simulation-based prediction. We present MASE-BDI, a multiagent system for environmental simulation that implements agents rationality through the belief-desire-intention mentalistic model. We implemented a semi-automatic integration with the software toolkit for uncertainty quantification PSUADE - Problem Solving environment for Uncertainty Analysis and Design Exploration. We propose some verification and validation and uncertainty quantification techniques to evaluate multiagent simulations and agent models. The application of this methodology in a land use case study is discussed and some results are presented. 

 

Profa Célia Ghedini Ralha (Este endereço de email está sendo protegido de spambots. Você precisa do JavaScript ativado para vê-lo.)

Coordenadora dos Seminários de Pós-Graduação em Informática 2017-2