Research On Distributed Security State Estimation Based On Wireless Sensor Network

With the development of sensor technology,communication technology and microprocessing technology,wireless sensor network is changing the way of interaction between human and nature,which affects human production and life.State estimation is very important for understanding the dynamic behavior of the system and performing some control tasks.Since sensors in distributed wireless sensor networks are physically exposed and communication networks are shared,attackers destroy the sensor networks in unpredictable ways such as covert,which increases the difficulty of designing distributed security estimation algorithms.Therefore,this paper studies the design of attack modeling,detection mechanism,defense strategy and distributed state estimation algorithm for the above problems existing in wireless sensor networks.Specific research contents are as follows:Firstly,when the physical layer of a system is subject to false data injection attack(FDI attack)and sensors occur fading measurements,an attack modeling,detection mechanism and distributed state estimator are studied.Considering that both FDI attack and fading measurements are random,a detection mechanism based on FDI attack observer is proposed.A distributed state estimator is designed based on minimum variance.Then,the stability of the proposed algorithm is analyzed.Secondly,a distributed consensus estimation problem is studied when FDI attack occurs in the sensor network channel.Considering that the communication channel between the sensor and the neighbor sensor is vulnerable to FDI attack,a residual-based detector is designed to detect whether the received information is attacked.Combined with the detection information,a distributed consensus estimator is designed by repeatedly fusing information matrix and vector of neighbors over wireless network.Then,a sufficient condition that the local estimation error is uniformly bounded in the mean square is proved.Finally,when denial-of-service Attack(Do S Attack)occurs in channels of sensor network and sensor occurs missing measurement,the problem of event-triggered distributed state estimation is studied.In order to reduce the influence of missing measurement,the local state estimation of each sensor is designed based on extended Kalman filter theory with compensation strategy.In order to reduce communication burden and guarantee the Gaussian performance of system,a random event-triggered mechanism is constructed.Based on local state estimation and event-triggered mechanism,a distributed state estimator is designed,which can defend against Do S attack.Then,stability of the design algorithm is proved.