A Fuzzy-Bayesian belief network approach to compute efficiency as a metric for IoT systems
摘要
In IoT applications, a multitude of devices communicate through interconnected networks, collaborating on decision-making tasks. Ensuring the efficiency of such IoT systems is crucial and requires addressing the management of resources. The article primarily focuses on the computation of efficiency in IoT applications, specifically considering metrics such as availability, functional correctness, and throughput. These metrics directly correlate with packet transfer information. Our main objective is to establish the efficiency of an IoT application by employing metrics related to packet transfer rate, utilizing a fuzzy-based Bayesian belief model. Key metrics such as availability, functional correctness, and throughput are commonly utilized to gauge the performance of packet transfer. To achieve this, we propose a Bayesian Belief Network (BBN) model that incorporates these metrics. To calculate efficiency, we have employed a fuzzy-based approach within the Fuzzy Inference System (FIS), which allows for efficient computation when crisp values of availability, functional correctness, and throughput are provided.