Generalized Multi-Sensor Kalman Filters with Detectable Measurement Delays and Their Comparison

Considering applications involved with remote sensing and estimation, this work investigates Kalman filtering of measurement data from multiple sensors with time-varying delays, intermittent measurements, and out-of-order arrival of the sampled data. Relying on time-stamping in the existing network protocols, it is assumed that the delay magnitudes are detectable. Two approaches are proposed to derive functionally equivalent Kalman filtering methods for optimal estimation in presence of the communication effects. An example is presented to verify the filtering performance and compare between the computational loads of the two methods under various conditions.