Research On Nodes Localization Technology And Localization Security For Wireless Sensor Networks

Wireless sensor networks are widely used for monitoring and controlling military,the environment,industry,agriculture and other fields due to their advantages of low power consumption,good fault tolerance and easy deployment on large scale.In the application ofwireless sensor networks,the position information of nodes plays a crucial role.Node positioning technology is one of the key technologies of wireless sensor networks.Wireless sensor network resources are limited,open deployment,unattended and other features make the node positioning process there is a security risk,so positioning security has great significance in wireless sensor network applications.After years of development of node localization technology,a series of node localization methods have been proposed and have achieved good results in certain applications.However,as an emerging research area,there are still more technical problems to be solved,such as:1)A variety of positioning algorithm systems are based on different assumptions and applications,using a specific method to solve the positioning problem.Now there is no universal optimal positioning algorithm system;2)High precision positioning algorithm relies too much on hardware support,and the calculation and communication costs are overhead,which is not conducive to the extensive application of the algorithm.3)Positioning algorithm System security is very fragile.The existing security technologies are usually based on encryption,ranging,angle measurement,strict time synchronization,or complicated statistical analysis,which contrary to the limited resources of wireless sensor networks,so positioning security algorithms still need to be further studied.In view of the above problems,this paper uses the method of theoretical analysis and experiment to deeply study node positioning technology and location security.The main research contents and innovations are as follows:(1)Since the cumulative error will seriously affect the localization accuracy in the process of node multi hop range-free localization,and the generation of cumulative errors is related to the error of the single hop distance,and the control of single-hop distance error can effectively reduce the cumulative error,the MLCEC algorithm based on cumulative error compensation has been proposed.Based on the quantitative statistical analysis of pre-order nodes on the shortest path,the algorithm modifies the hops of the shortest path from the beacon node to the unknown node,and puts forward the concept of k-hop average hop distance,at the same time,combined with the hop correction,the cumulative error has been better controlled,greatly reducing the localization error.In allusion to the abnormal distribution of beacon nodes,a simple detection their Collinearity method is proposed to improve the robustness of the localization algorithm.Experiments and simulation results show that the extra overhead of the algorithm is small and it can achieve a better localization performance when the ratio of the beacon nodes is small.The algorithm has good expansibility.(2)In order to overcome the problem of anisotropy due to non line of sight,density and other causes,MLKT localization methods based on kernel Tikhonov regularization was presented.This method constructs a mapping function by kernel Tikhonov regularization between hop-counts and physical distances for pairs of anchors,and then obtains the no-anchors to anchors estimated distance using the measurement of hop-counts.The theoretical analysis and simulation results show that the improved method can solve ill-condition problem and and improve the positioning accuracy with avoiding the influence of anisotropy,and has the less calculation cost and parameters,thus it is suitable for the harsh nodes deployment environment with high practical value.(3)By the analysis of the exposed attack model,for the characteristic that the probability of the attack node being the unique preorder node(UPN)on the shortest path between node communications greatly increases,analyze the quantity standard variance of UPN,and put forward UPNDA detection algorithm.The calculation of the algorithm is simple and can effectively detect and prevent the exposed attacks.What's more,it can reduce the effect of attacks on localization algorithm.In view of the influence of the hidden attack on the topological structure of the network,analyzing the anomalous change of the before and after attack of the neighbor nodes,the NSDA detection method based on neighbor node set analysis is proposed and the attacked nodes are classified and processed,minimizing the attack effect as far as possible.In the case of large communication range of the attack node,the WRDA method based on the ring of wormhole is proposed.Through the probability analysis of node connectivity,based on some of the characteristics of the attacked network like heterogeneous networks,the algorithm constructs the wormhole ring and normal node ring for the nodes within the attack range,and using the linear regression analysis method,through the normal node rings of the wormhole ring edge to reduce the influence of the attack on localization.