Mobile user location determination using extreme learning machine

Abstract

There has been a rapid convergence to location based services for better resources management. This is made possible by rapid development and lower cost of mobile and handheld devices. Due to this widespread usage however, localization and positioning systems, especially indoor, have become increasingly important for resources management. This requires information devices to have context awareness and determination of current location of the users to adequately respond to the need at the time. There have been various approaches to location positioning to further improve mobile user location accuracy. In this work, we examine the location determination techniques by attempting to determine the location of mobile users taking advantage of signal strength (SS) and signal quality (SQ) history data and modeling the locations using extreme learning machine algorithm (ELM). The empirical results show that the proposed model based on the extreme learning algorithm outperforms k-Nearest Neighbor approaches.