On optimal state estimation with multiple packet dropouts

Due to unreliable communication between local sensors and the processing center, packet dropouts may happen during transmission. Two existing methods for linear minimum mean-squared error (LMMSE) estimation with multiple packet dropouts were obtained completely or partially based on a stochastic parameter system constructed by augmenting the original state and measurement. They have a high computational load, unclear measurement residual characterization and tough requirements on initialization. To overcome these, an alternative form of LMMSE estimation with multiple packet dropouts is derived. Under a Gaussian assumption, the minimum mean-squared error (MMSE) estimation with multiple packet dropouts is also derived. Numerical examples are provided to compare performance of the proposed estimators.