Target Localization in CEMS Based on Shunt-Wound Radial Basis Function Network

摘要

In the complex electromagnetic space, the traditional methods can not normally work, because of the absence of the model information about the space. Due to the rapid development of artificial neural network, the data-driven method is available and effective for solving the localization problems in CEMS (complex electromagnetic space). However, because the unknown electromagnetic environment can not support us to select the input of training data, the inputs of training data are nonuniform, which would result in the unexpected performance of RBF (radial basis function) network. For solving the problem, the convex transformation and the shunt-wound structure are proposed and employed in this article. Furthermore, the experimental results show the proposed RBF network is better than the single RBF network.