Adaptive event-triggered distributed estimation over wireless sensor networks with uncertain communication topologies

This study investigates an adaptive event-triggered approach for consistent estimation in discrete time-varying systems that operate within sensor networks characterised by dynamic and unpredictable communication topologies. A Markovian stochastic process governs topology switching, and complete knowledge of the transition probabilities of transitioning from one topology to another is not fully known. To optimise the energy consumption of sensor networks, a novel adaptive event-triggered mechanism (AETM) was developed to regulate data transmission efficiently. The performance of the proposed AETM is further enhanced by incorporating a historical data smoothing technique, which mitigates the effects of device jitter and significant stochastic noise, thereby minimising unnecessary triggering events. The desired parameters for the distributed estimator are then determined by iteratively solving a set of inequalities that guarantee the mean square exponential stability and l2-l∞ robustness of the filter error system. Finally, the effectiveness of the proposed approach is demonstrated through a numerical simulation example.