Adaptive Periodic Communication over MQTT for Large-Scale Cyber-Physical Systems

The periodic data acquisition is essential to accurate monitoring in the control loop of cyber-physical systems (CPS). However, providing periodic communication for large-scale CPS is a challenging issue, because the network resources are shared by a large number of nodes. Consequently, it usually happens that, though the sender periodically sends messages, the arrival intervals of messages on the receiver are larger than the period. In this paper, we suggest an adaptation framework for periodic N-to-1 communication over Message Queue Telemetry Transport (MQTT), a well-known publish-subscribe (pub-sub) messaging protocol. Our suggested framework monitors the usage of computation and network resources, and adaptively adjusts the publication period in consideration of characteristics of the pub-sub protocol, such as split TCP connections and QoS levels. The experiment results clearly show that the adaptation framework can provide superior timeliness of messages.