Temporal-spatial planning of communications for networked state estimation
Chao Yang · 2013
This thesis investigates the issue of communication planning in networked control systems, with the objective to improve state estimation quality subject to communication resource constraint. It considers both temporal and spatial planning of the communication. The temporal planning of communication refers to the transmission time scheduling between a sensor and a remote state estimator. The thesis studies two problems of temporal planning. Both of them consider state estimation of a linear time-invariant system, in which a sensor takes measurement of the system state and transmits it to a remote estimator via a communication channel, and the remote estimator computes an estimate of the system state. The first one studies offline communication scheduling of the sensor and provides the optimal schedule; while the second considers hybrid scheduling, which makes use of online information based on the offline scheduling, and proposes several novel algorithms of hybrid scheduling. Spatial planning refers to the design of communication topology for a group of sensors. The thesis studies the problem of estimating the state of a system measured by multiple sensors. Optimal sensor communication topology designs are provided for the homogeneous and heterogeneous sensor networks, respectively.