Cluster and grid based weighted K-Nearest Neighours for outdoor location estimation

This paper describes an approach to accurate mobile user location estimation outdoors without GPS or triangulation. The partitioning of the wireless environment into clusters that allow accurate modeling of the propagation environment and grid based approaches is compared. The models are constructed from historical data obtained by monitoring of the received signal strengths (RSSs) from the mobile stations (MSs). The performance is evaluated using real data in a GSM network around the Queen Mary campus.