WLAN Indoor GA-ANN Positioning Algorithm via Regularity Encoding Optimization

To begin with, for indoor location system, the necessity of research on genetic neural network and its math model are introduced. Then, by analyzing principle of genetic optimized artificial neural network, an indoor location math model of genetic neural network is established. As for various coding types, regularity is taken as the measurement to determine the best coding type for parameter optimization. By analyzing theory of splicing/decomposable coding, the advantages of regularity for such coding type are proved. Finally, through simulation comparisons, to select a regularity coding type for GA-ANN can improve positioning accuracy for indoor environment effectively.