Event-Triggered Transmission Policies for Harvesting Powered Sensors With Time-Varying Models

We consider a wireless control system where sensors transmit their measurements to a controller over a fading channel. We assume that each sensor is equipped with a rechargeable battery and can harvest energy from the environment or other energy sources. To predict the harvested energy, we consider a stationary Markovian model conditioned on a scenario and estimate the unknown parameters based on empirical measurements. The transmission energy is assumed to be inversely proportional to the communication channel gain, which is described by a compound distribution. To increase energy efficiency, we use a continuous event-trigger allowing the sensors to transmit new data only when certain conditions are satisfied. Finally, we analyze the exponential stability of the closed-loop nonlinear system with multiple sector-bounded nonlinearities based on the input-delay method and Lyapunov–Krasovskii technique.