Event-based Colored-Noise Kalman Filtering for Improved Resource Efficiency
Eva Julia Schmitt, Benjamin Noack · 2023
In modern automated systems, the number of agents and sensors is rapidly increasing and with them the energy consumption and the burden on communications networks. One way to increase the energy and spectral efficiency in sensor networks is to replace periodic transmissions between nodes with event-based transmissions and to use matching estimation techniques at the receiving nodes. Since multi-sensor systems and smart event-trigger designs often lead to correlations between measurements due to correlated process and measurement noise, estimators that can handle such correlations are required to guarantee good performance. In this paper, an event-based colored-noise Kalman filter was developed and specifically designed for an finite impulse response-based stochastic trigger. Nevertheless, the concept is suitable for a wide class of correlated input data.