Research On Localization Technique Of Wireless Sensor Network In Constrained Space

With the development of information technology, the requirement of people to obtain informationfrom the objective physical world gradually strengthened, the sensing technology is applied in manydifferent scenarios. As a result of it, wireless sensor network, as the central of informationinterconnection, is widely used in different fields such as environment monitoring and nationaldefence. In wireless sensor network, localization for node is very important, which leads to a fact thathigh localization accuracy is helpful for detecting the position coordinates of an event or a node and itis the foundation of the node deployment, data aggregation and security. So it is particularlyimportmant to solve localization of wireless sensor network.Most of the existing location technology confined to the outdoor environment, but more and moreconstrained space gained the requirement of localization, such as coal mine tunnels, archaeologicalsite and indoor corridors. Constrained space is limited by such factors as multipath, changinginterferences and non-line-of-sight as well as linear-structure network topology, which leads to theexisting location technology can not satisfy the location demands. In this thesis, we proposedlocalization algorithms for constrained space around factors as localization accuracy, security and lowenergy consumption. The mian research work and contributions are summarized as follows:(1) A gradient-based security location algorithm is proposed in this thesis, in which gradient isbrought based on hyperbolic location model. This algorithm lowers the computational complexity sothat it can satisfy the environmental demands of constrained space and achieve the high-accuracylocalization. Meanwhile, to improve the security of the algorithm, the resistant ability is enhanced andthe selective pruning process for inconsistency measuring is also adopted. We finally improved thelocation accuracy by using main frequency strategy.(2) A asymmetric event-driven localization algorithm is proposed in this thesis, which is based onthe combination of event distribution and anchor node achieve a distributed location estimationstrategy, so that it can satisfy the localization requirement of constrained space and achieve thehigh-accuracy localization with a small amount of events and anchor nodes. Meanwhile, to improvethe accuracy of the algorithm, a set of candidate events are adopted to prune the event which does notmeet with location accuracy requirements.(3) Baesd on the existing wireless sensor network node and the virtual machine platform Agillawith multi-Agent, we designed and achieved a system for localization of wireless sensor network inconstrained space. The system can flexibly executes the localizaiton commands, what’s more, it has a certain expand ability and adaptive ability. For the localization system designed, we made a detailedintroduction on the system architecture and the design and realization of the various modules in thisthesis. In addition, combined the two methods put forward in this thesis, we explained the practicalapplication of the system.