Indoor Positioning Algorithm Based on Parallel Multilayer Neural Network

Abstract

With the growing use of wireless devices and easy access to received signal strength (RSS), there are increasing attentions on utilizing RSS technology for indoor positioning. In this paper, we propose indoor positioning algorithm based on parallel multilayer neural network (PMNN) for RSS-positioning. Two multilayer neural networks are trained separately to estimate x and y coordinate index, using RSS data gather at the target node from access points (APs). Eachmultilayer neural network structure integratesdenoising section and positioning section into a deep structure. Experiment results demonstrate the feasibility and suitability of the proposed algorithm.

6 Figures and Tables

Cite this paper

@article{Dai2016IndoorPA, title={Indoor Positioning Algorithm Based on Parallel Multilayer Neural Network}, author={Huan Dai and Hong-bo Liu and Xiao-Shuang Xing and Yong Jin}, journal={2016 International Conference on Information System and Artificial Intelligence (ISAI)}, year={2016}, pages={356-360} }