Multi-objective Filtering for Discrete-time Systems in the Presence of Data Packet Drops

This paper studies the multi-objective filter design for linear time-invariant discrete-time systems over communication networks, where data packet dropout occurs from the observation side to the filter. The desired filter is able to achieve the optimal estimation error covariance under white Gaussian noises, together with a guaranteed worst case performance against disturbance/model uncertainty. By game theoretical approach, a bounded Nash equilibrium strategy is analytically conducted in terms of two cross-coupled modified algebraic Riccati equations. Moreover, a numerical example is included to show the validity of the current results.