Modelagem e simulação de redes bayesianas para o cálculo de probabilidade de falha em sistemas IoT na saúde

Abstract

The internet of things has been applied in several contexts: from smart cities, education, supply chain and health. The implantation of this technology provides benefits to life, such as: remote control of pests in agriculture, monitoring of the supply chain, improvement in the physical and virtual environment in education and monitoring of patients. However, despite the benefits, there are challenges embedded with the implementation of this technology, among which stand out maintaining data privacy and security, ensuring data integrity and reliability, as well as energy cost management. In particular, with regard to data privacy and security, as it is one of the biggest challenges in the area, it is necessary to evaluate the probability of the components failing and, consequently, causing this problem. It is in this context that this work proposes to identify, modelling and calculate probability of failure, using Bayesian Networks. The models built from this technique allow estimating different scenarios in the use of an Internet of Things network. The methodology used was the mixed approach, combining characteristics of the qualitative and quantitative approach when carrying out a systematic review of the literature, application of forms to collect the perception of specialists, in addition to the use of other techniques to strengthen the results, among which Delphi stands out. and Noisy-OR. And the results showed that through the use of the model it is possible to evaluate different scenarios for the use of Internet of Things networks, as well as simulating the effect of probability of failure on the critical components of the system.