Distributed Estimation Of Sensor Networks Under Wireless Channel Attacks

Wireless sensors been widely applied in various industries in social life due to advantages such as small size,low cost,self-organization and abilities including perceiving,processing and transmitting data.Wireless sensor networks can be found in many fields like important military with significant secrets and national defense,home automation,and health care which are closely related to our lives.In order to decentralize data processing,distributed estimation algorithms over sensor networks and data fusion have a rapid development which are the hotspots of research at the present stage about extensive uses of positioning,monitoring and tracking targets.However,with the rapid development of network technology,wireless sensor networks are increasingly facing with various problems of network attack which can seriously threaten the security and stability of system.Considering attacks of sensor networks,this paper studies the distributed security estimation problems under two different channel attacks.The main results are as follows:Resolving problems of energy limitation of sensor in wireless sensor networks and false data injection attacks in network channels,distributed security estimation algorithm is proposed in which a novel event-triggered mechanism is designed and an attack detection strategy is constructed with the use of Chi-square.Besides,data fusion is also applied in the algorithm.Then,the boundedness of transmission data information matrix and the convergence of mean square error can be proved theoretically and the related parameters are given in the end.Finally,advantages of the algorithm are verified and the effectiveness and accuracy of the attack detection are shown by a numerical simulation.Resolving problems of energy limitation of sensor in wireless sensor networks and denial of service attacks in network channels,distributed security estimation algorithm where a novel eventtriggered mechanism based on the acknowledge character(ACK)of communication feedback data packet is designed and a strategy of data compensation is demonstrated,is proposed with using data fusion.In addition,the boundedness of transmission data information matrix and the convergence of mean square error can be confirmed with theory and the parameters relating to algorithm are given.Finally,a simulation is given to illustrate the effectiveness of the algorithm and the superiority of the system performance.