Cloud-based Self-Organizing Localization for wireless sensor networks in mixture environments of LOS and NLOS

Wireless sensor networks (WSNs) are an essential technology for Internet of Things (IoT) and Machine-to-Machime (M2M), which attempt to accommodate physical things on the Internet. WSNs are presumably applied in environments where a diverse space contains obstacles. In radio propagation, the space is the mixture space of Line-of-Sight (LOS) and Non-Line-of-Sight (NLOS), and the WSN topology is anisotropic. We previously proposed a Self-Organizing Localization (SOL), which is a localization applied to Self-Organizing Maps. A SOL is capable of estimating accurate node location with only neighbor topology information. However, it suffers from the following two issues: misestimation occasionally occurs in the LOS/NLOS mixture space and the amount of inter-node communication increases owing to the iteration exchanges of node locations between nodes. In this paper, we propose a cloud computing-based SOL that solves the above issues.