المجلة الدولية للعلوم والتقنية

Smart building systems in major cities are equipped with Internet of Things (IoTs) devices to monitor various indicators within these buildings. To avoid service quality degradation, performance indicators related to these systems must be identified before building implementation. Through these indicators, the system's behavior can be understood and analyzed quantitatively performance estimation. In this study, a queue network model was used to model and simulate a smart building consisting of two floors and five rooms. The main room is equipped with five sensors. The system was simulated using the Java Modeling Tool, enabling the estimation of performance indicators and the examination of various scenarios. These scenarios involved the precise estimation of parameters for each scenario based on previous studies, including scenarios that explored changes in the number of servers.The following indicators were adopted in this study to analyze the system's behavior such as average response time, system message count, resource utilization ratio, flow rate, and message loss rate. Simulation results indicated that using 8 cores resulted in very high execution speed, cost, and energy, with optimal performance indicators such as a significantly reduced average response time by 73 ms for the number of servers.Analyses conducted at three levels revealed that the bottleneck occurs at the edge computing level, where the utilization rate is 100%. The results also demonstrated that the proposed queue network model is a valuable tool for studying quality of service requirements for Internet of Things systems. Additionally, the analyses derived from the simulation can assist designers of smart building systems in enhancing the performance of these systems before implementation. Keywords: Internet of Things, smart building, performance indicators, queuing model.