Research And Improvement On Localization Algorithm In Wireless Sensor Networks

With the speedy development of microelectronic devices, wireless network technology and sensor Technology, research environment of wireless sensor networks has also been improved, and its equipment costs has been decreased substantially, wireless sensor network has been applied to many areas in people’s life. Localization technology is one of crucial technologies in the wireless sensor networks, and is the foundation of achieving many applications of wireless sensor networks. By researching the algorithm of localization, especially the high-precision, low-power and low-cost localization algorithm, it could help us to develop the WSNs’ ability of solving practical problems, and improve people’s quality of life.Firstly, the paper simply introduced relevant content of WSN, such as WSN’s research background, research status at home and abroad, whole network characteristics, single component structure, and common crucial technologies. Secondly, the paper explained the localization technologies in the crucial technologies, starting from the basic terms of positioning, introducing position classification under different kinds of reference index, and proposing evaluation of algorithm performance. Then the next two chapters elaborated, deducted and simulated the main research contents of this paper:DV-Hop and Convex in detail, including the principle of algorithms, existing improved algorithms and ones in this paper. Finally, the last chapter made a conclusion and prospect on the whole paper.There were mainly two innovative parts in this paper:on one hand, improved weighted DV-Hop algorithm based on least squares optimization was proposed on the basis of existing one, in this part, concept of GPS localization error was put forward in order to measure the stability of the algorithm; least squares was used to optimize the whole network’s average one-hop distance of beacon nodes; hyperbolic algorithm rather than maximum likelihood estimation was used to estimate the coordinates, and these changes made the algorithm keep a better positioning effect in the case of that coordinates of beacon nodes existed error. On the other, convex problem in the localization algorithm was replaced by optimal solution of multivariate equation solved by PSO; fitness function was optimized by weighted factor that was inversely proportional to hop; choose some optimal beacon nodes to locate in the iterative, and make sure the localization had a higher accuracy on the condition of unevenly distributed beacon nodes equally.