Bringing MQTT Brokers to the Edge: A Preliminary Evaluation

An event‐driven fusion framework with auto‐scaling of edge intelligence for resilient smart applications in developing countries

“…To address the above objectives, an event‐driven information fusion framework is proposed, that runs using docker containerization technology on the fog‐cloud architecture, to provide resilience to the application and auto‐scale edge intelligence according to the change of environment state. Moreover, the framework uses Message Queuing Telemetry Transfer (MQTT) protocol 7 for event‐driven data communication between docker microservices in the fog node as well as with the cloud, as it is light‐weight and provides reliable communication in case of limited network. Overall, the contributions of this work are as follows: An event‐driven information fusion framework is proposed with a stand‐alone fog node and a cloud for a low‐cost IoT application, deployed for monitoring rural and semi‐urban locations of developing countries, suffering from limited network connectivity and frequent power cuts. The fog node runs an event‐driven controller, to provide resilience to the application, detecting intermittent connectivity issues and frequent power outages and re‐configuring data processing on fog‐cloud architecture in real‐time. An efficient information fusion model is proposed for providing edge intelligence in the presence of limited network connectivity and power outages. State change of the monitored environment with respect to dynamic population rush is modeled using queuing theory and a novel auto‐scaling approach is proposed using combination of statistical and probability based estimation algorithms for lightweight, resource efficient, and timely service on the edge device. …”