Distributed H_? Filtering For Wireless Sensor Networked Systems With Switching Topology

With the rapid development of electronic and wireless communication technology,wireless sensor networks(WSNs)have been widely used in many fields.Numerous engineering problems can be modeled as the signal estimation problems of dynamic systems.Therefor designing a high reliability distributed filtering algorithm has been a hot research topic WSN.A key concern of distributed filtering is how to model the uncertainty of data acquisition by sensor nodes and to study the effect of distributed network topology on estimation performance.Considering the characteristics of wireless sensor networks,the design problems of distributed H? filtering with switching topology and event-triggered transmission schemes are studied in this paper.The main work are stated as follows:(1)For continuous linear systems,the problem of distributed filtering over sensor networks with switching topology and topology-dependent event-triggered transmission schemes is studied.First,a new monitoring mechanism is proposed to identify the topology switching time,and it is assumed that the system output is sampled by sensors when the network topology switches,respectively.Based on the time-varying sampled-outputs,a group of distributed topology-dependent event-triggered transmission schemes is introduced to reduce the amount of transmitting sampled-data and energy consumption.Under a unified framework of switching topology and event-triggered transmission scheme and the network-induced delays,the networked filter error system is modeled as a switched time-delay system.By using a Lyapunov-Krasovskii functional,a bounded real lemma is obtained such that the filtering error system is exponentially stable with a prescribed attenuation level,and a co-design method on the networked filter gains and the event-triggered parameters is also proposed.(2)For discrete linear systems,the problem of distributed filtering for sensor networks with missing data and switching network topology is studied.First,a redundant channel is introduced to reduce the possibility of packet loss in the case of data transmission through a single channel.A set of Bernoulli random sequences is employed to model the phenomena of communication link failure both the main channel and the redundant channel.In order to ensure that the distributed filters can follow the changes of the network topology,a switching signal is introduced and the switching of several topological modes is described by a Markov chain.The filtering error system is modeled as a stochastic time-delay system.By using a Lyapunov-Krasovskii functional,a bounded real lemma is obtained such that the filtering error system is exponentially stable with a prescribed attenuation level.Finally,an illustrative examples is given to show the effectiveness of the designed filter.