Event-Driven Offloading and Load-Balancing With Quantized Control for Fog-Enabled UAVs

This article presents the development of a quantized control framework for a distributed fog computing network (DFCN) for event-driven offloading and load-balancing to process the visual information referred to as vision data.The DFCN consists of a network of scalable fog nodes (FNs) deployed to infer vision data in a given time bound called the service time. At first, the dynamic model of a DFCN is developed using the system identification techniques. Later, the control design for stability and desirable transient response with event triggering, including the effects of quantization at the input and the output of the DFCN, are presented. Finally, we validate the proposed theory for processing vision data captured by the unmanned aerial vehicle under a given service time by auto-scaling the FNs in a DFCN. Simulation and real-time experiments validate the efficacy of the proposed control design.