Sensor scheduling for distributed filtering under limited communication resources

In this paper we consider the problem of sensor scheduling for distributed estimation with power constraint. To satisfy the requirement of energy consumption, we assume that each sensor sends data to its adjacent sensors with a given probability, where the state of the system is estimated by means of distributed estimators. First, we design optimal estimator gain to minimize the estimation error covariance for each sensor, and then we find an upper bound of the expected state error covariance, provide a sufficient condition to guarantee the stability of the proposed estimator. To find the optimal communication probability, we further formulate the sensor scheduling problem as an optimization problem, which is relaxed to a convex optimization in the form of LMIs.