Stochastic Sensor Scheduling for Centralized State Estimation

A stochastic sensor scheduling problem is studied in this paper. A group of sensors take measurements of the states of a process and send the data to a remote estimator. Each sensor communicates with the estimator randomly and the communication is determined by a Bernoulli random variable. Meanwhile, the sensors have limited power resources for communication. The design of the communication rates which optimizes the estimation performance under the limited communication resources is studied in this paper. This optimization problem is relaxed and solved by tractable numerical algorithms. Examples show that satisfactory performance is obtained.