Modeling and Simulating Internet-of-Things Systems: A Hybrid Agent-Oriented Approach

Abstract

The focus of the Internet has recently shifted from current computers and mobile devices to everyday objects, people, and places; consequently, the Internet of Things (IoT) promises to be not only a compelling vision but the actual driving force of the upcoming fourth Industrial Revolution. Novel cyber-physical, customized, and highly pervasive services are impacting our lives, involving several stakeholders and fostering an unseen globally interconnected ecosystem. However, IoT system development is a multifaceted process that's complex, error-prone, and time-consuming. Although modeling and simulation are crucial aspects that could effectively support IoT system development, an integrated approach synergistically providing both of them is lacking. The authors propose a hybrid approach that uses agents for IoT modeling and OMNeT for simulation, providing mapping guidelines between the agent paradigm and the OMNeT simulator abstractions. The proposed approach has been applied in small-, medium-, and large-scale IoT scenarios, where relevant performance indexes of IoT entities communication have been measured and analyzed.