期刊名称：International Journal of Innovative Research in Computer and Communication Engineering
印刷版ISSN：2320-9798
电子版ISSN：2320-9801
出版年度：2014
卷号：2
期号：4
出版社：S&S Publications
摘要：Wirel ess sensor network node localization is an important research area that creating more innovations in the field of WSN. The exact position of each node is necessary for efficient routing of packets and location - aware services. Existing systems are not produ cing satisfactory results. In this paper proffers the use of time difference of arrivals (TDOA) information with neural network to e stimate the node location. Two artificial neural net work models - Radial Bases Function Network and multilayer perceptron mod els are used for wireless sensor network node localization problem. Time difference of arrival data is used to calculate the distance between anchor nodes to sensor nodes, this information data is used to train and test the neural network models. Simulatio n result gives the pe rformance Radial Basis Function Network and Multi layer perceptron Network in terms of root mean square error by varying training data density.
