Reliability Modeling of Dynamic Spatiotemporal IoT Considering Autonomy and Cooperativity Based on Multiagent

Reliability modeling in the context of the Internet of Things (IoT) has received considerable attention from researchers to ensure the safety and stable operation of the relevant systems. However, current methods of reliability modeling are incapable of adequately characterizing the autonomy and cooperativity of the Dynamic Spatiotemporal IoT (DSIoT). To solve this problem, this study proposes an agent-based method of reliability modeling. The authors discuss the autonomy and cooperativity of the DSIoT in case of failure and summarize 11 typical failure states of its functional components. We then develop a multiagent-based framework for the reliability modeling of the DSIoT. Methods of failure modeling are analyzed to describe the propagation of failure within a node and among multiple nodes. Two kinds of criteria of failure are proposed for IoT systems with dynamic spatiotemporal characteristics, and a Monte Carlo simulation-based approach is given to evaluate the reliability of the DSIoT. We also illustrate the uses and effectiveness of the proposed method by using a case study involving a swarm of 16 intelligent unmanned aerial vehicles.