Cluster-Based RF Fingerprint Positioning Using LTE and WLAN Outdoor Signals

摘要

In this paper we evaluate user-equipment (UE)positioning performance of three cluster-based RFfingerprinting methods using LTE and WLAN signals.Real-life LTE and WLAN data were collected for theevaluation purpose using consumer cellular-mobilehandset utilizing ‘Nemo Handy’ drive test software tool.Test results of cluster-based methods were compared tothe conventional grid-based RF fingerprinting. Thecluster-based methods do not require grid-cell layout andtraining signature formation as compared to the gridbasedmethod. They utilize LTE cell-ID searchingtechnique to reduce the search space for clusteringoperation. Thus UE position estimation is done in shorttime with less computational cost. Among the cluster-basedmethods Agglomerative Hierarchical Cluster based RFfingerprinting provided best positioning accuracy using asingle LTE and six WLAN signal strengths. This methodshowed an improvement of 42.3 % and 39.8 % in the 68thpercentile and 95th percentile of positioning error (PE)over the grid-based RF fingerprinting.