| With the rapid development of Wireless Sensor Networks， WSNs has become a newtechnology to access to information. Wireless Sensor Networks consist of a self-organizingnetwork which makes up with a lot of low-cost tiny sensor nodes that can perform sensing,computing and communicating. The nodes in WSN can work in cooperation to transmit theprocessed information to the remote monitor. As one of the key technologies of WSN, localizationtechnology has great effect to improve the performance and value of WSN. Nowadays,positioning technology in WSN has become a research hotspot. There are many localizationalgorithms proposed by researchers, but most of the algorithms without extra equipment are lowaccurate. According to the problems in the algorithms, the foundational technology includes thestructure and particularity of wireless sensor networks is discussed in this thesis. Then this thesisintroduces and analyzes the advantages and disadvantages of several classical algorithms ofrange-free and range-based technology. This paper proposes two algorithms which have betterlocalization accuracy.Two localization algorithms are proposed in this Paper, one is nonlinear weighted least squarealgorithm，the another is improved APIT(Approximate Point-in-triangulation Test) algorithm.The first algorithm proposes a modification method of gauss filter to improve the accuracy ofRSSI ranging on the wireless channel model. Then we put forward a nonlinear weighted leastsquare method and use Taylor series to reduce the computational complexity, then we achieve anode localization algorithm with higher accuracy. The improved APIT algorithm improve thePoint-in-triangulation Test and the localization of undefined nodes in traditional APIT algorithmbase on the discussion of the principle and shortcomings in traditional APIT algorithm. Thesimulation of the two proposed algorithms is made at the last part of the paper and the simulationanalyses prove the better localization accuracy. |
PDF Full Text Request