H∞ filtering for networked systems with uncertain packet dropouts and quantization

This paper studies the H∞ filtering problem for networked systems subjected to random packet dropout and quantization effects. The estimated packet dropout probability with error bound is considered here due to the observation that the exact one is difficult to obtain in practical systems. The packet dropout process is modeled as a Bernoulli sequence and the quantization is modeled into system uncertainty by the sector bound approach. The filtering error system is then obtained as a stochastic system with both system and probability uncertainties. A filter design procedure is given based on sufficient conditions for the filtering error system to be stochastically stable with an H∞ performance. Finally, an example is given to illustrate the effectiveness of the proposed method.