Distributed Event-Triggered Set-Membership Filtering Subject To Cyber Attacks

Distributed filtering based on wireless sensor networks is a hot research field.In engineering practice,wireless sensor networks are usually subject to energy constraints and malicious cyber attacks.As such,this paper is concerned with the distributed set-membership filtering problems via wireless sensor networks subject to unknown-butbounded noises with consideration of both security and energy-saying.Several kinds of common network attacks are considered,and their mathematical models are established for the analysis and design of the distributed filter.At the same time,the event-triggered communication protocol is designed to lower the communication frequency by reducing unnecessary signal transmission between nodes,so as to improve the efficiency of energy utilization.Then,the sufficient conditions for the existence of the expected distributed set-membership filter are obtained and the filter design algorithms are proposed.Furthermore,the locally optimal filter gains are found via solving optimization problems.Specifically,the research contents of this paper include the following aspects:· Considering the wireless sensor networks with sensor saturation and quantization effect,the distributed set-membership filtering problem is investigated for a class of discrete time-varying systems subject to replay attack.With the help of linear matrix inequality technique,the solvability problem of the distributed set-membership filter design is transformed into the feasibility problem of a group of recursive linear matrix inequalities.The desired filter gains are obtained by solving the corresponding recursive linear matrix inequalities.Furthermore,a convex optimization problem is solved to find the locally optimal filter.· The problem of distributed set-membership filtering is studied for a class of discrete time-varying systems via wireless sensor networks using event-triggered communication protocol.The influence of event triggering communication mechanism on measurement transmission is analyzed by using analytic geometry method.Furthermore,considering the measurement of the node itself and the measurement of the neighboring nodes scheduled by the event-triggered protocol,a minimum volume ellipsoidal distributed set-membership filtering algorithm is proposed.Simulation results show the effectiveness of the proposed algorithm.· Under the scheduling of coding-decoding communication strategy,the distributed set-membership filtering problem is solved for a class of discrete time-varying systems over digital wireless sensor networks.Considering the limited energy and network bandwidth,a dynamic coding-decoding communication strategy is designed.By using recursive linear matrix inequality technique,sufficient conditions for the existence of the filter with desired filtering performance are obtained.An optimization algorithm is proposed to seek the gain of the filter which can guarantee the locally optimal filtering performance.· Considering the wireless sensor networks with channel fading,study the distributed set-membership filtering for general nonlinear systems under bias injection attack.The measurement signal transmitted in wireless channel is affected by both bias injection attack and channel fading.By employing Taylor formula,set-membership filtering and linear matrix inequality techniques,sufficient conditions for the existence of the expected distributed filter are obtained.The exact filter gains can be obtained by solving a set of recursive linear matrix inequalities.Furthermore,an optimization algorithm is proposed to look for the locally optimal filter.Simulation results show the effectiveness of the proposed filtering algorithm.· For wireless sensor networks subject to hybrid network attacks and event-triggered protocol,the distributed probability-guaranteed set-membership filter is designed for a class of stochastic nonlinear systems.The random hybrid attack strategy includes two kinds of most common attack behaviors,namely deception attack and denial-of-service attack.The concept of probability-guaranteed set-membership filtering is introduced to deal with the random characteristics of attacks.By means of linear matrix inequality technique and stochastic analysis method,the sufficient conditions for the existence of the distributed filter with given filtering performance are obtained.Furthermore,two optimization algorithms are proposed to design locally optimal filters.Finally,a simulation example is given to verify the effectiveness of the proposed distributed secure filtering algorithm.