Quantized state estimation for sensor neterworks

Limited bandwidth is an unavoidable constraint for data transmission from local sensors to the fusion center over sensor networks. For a networked target tracking system with single sensor, which consists of a state-vector and an observation-vector, the quantized state estimation with bandwidth constrained is studied under the centralized frame for wireless sensor network system in this paper. Firstly, the augmented state method is adopted to provide a more effective frame to solve the quantized filtering. Namely, the quantized error vector can be taken as a state component by the augmented technology. Accordingly, a novel quantized state estimator is proposed by using Kalman filter (KF). Secondly, because Strong Tracking Filtering (STF) has better estimation accuracy and the ability of processing state abrupt change than Kalman filter for the case with uncertain state model, another novel quantized state estimator is proposed by combing the STF with augmented state model. Finally, two examples are demonstrated to show the validity and superiority of the proposed quantized estimators.