Journal of Industrial and Management Optimization · 2005 · 22 citations · 0 references
Natural SciencesMathematicsSystems EngineeringOptimizationSchedulingStochastic ProcessesOptimal ControlSensor NetworksSensor NoiseStochastic ControlStochastic OptimizationUncertainty EstimationEstimate VarianceFilteringFiltering TheoryMinimum Variance Unbiased EstimationOptimal FilteringScheduling Under UncertaintySensor SchedulingStochastic Optimal Control
In this paper we consider the question of sensor scheduling indiscrete time. The basic problem is to design a linear filterwhose output provides an unbiased minimum variance estimate of asignal process whose noisy measurements from multiple sensors areavailable for input to the filter. The problem is to select onesource (sensor data) dynamically so as to minimize estimationerrors. We formulate the problem as an optimal control problem. Byanalyzing the positive semi-definite property of the errorcovariance matrix, we develop a branch and bound method tocalculate the optimal scheduling strategy and give a numericalresult for interpretation.