Self-organizing map approach for determining mobile user location using IEEE 802.11 signals

This paper describes a study towards user location system based on Kohonen Self-Organizing Maps (SOM) algorithm for locating user location with the existing Wi-Fi signals. With the novel use of Wi-Fi based on the IEEE 802.11 standards: inferring the location of a wireless client from signal quality measures, the Kohonen SOM paradigms were used. SOM as an unsupervised learning techniques of Artificial Neural Network (ANN) capable for summarizing high-dimensional data which cause region of the network to respond similarly to certain input patterns by analyzing the signal strength or signal-to-noise (SNR) of the wireless access points (AP) that enable a wireless networked device to infer the location of wireless client. Location estimation is then computed using SOM on sample sets. We outline the study leading to the system and provide location performance metrics where a user who is using wireless local area network (WLAN) will be able to detect with the existing of Wi-Fi signal nodes with the SOM technique, highlighting position at the current time.