Positioning in heterogeneous small cell networks using MDT RF fingerprints

This paper presents a performance evaluation of Radio Frequency (RF) fingerprinting framework in heterogeneous Long Term Evolution (LTE) networks using Minimization of Drive Testing (MDT) measurements which allow automated construction of extensive RF fingerprint training databases. Fingerprint positioning was studied in heterogeneous small cell and regular macro networks using LTE cell detection performance requirements. Goal was to study how big effect the cell detection requirements have on positioning accuracy in interference limited inter-frequency small cell network deployment. The results show that RF fingerprint performance suffers degraded cell detection performance in fully loaded network deployments and results in lower dimensions of the RF signatures. However, in denser network deployments, the absolute positioning errors remain good being below the E911 emergency positioning requirements. This suggests that in dense small cell networks, MDT training databases can provide a good basis for location assisted Radio Resource Management (RRM) algorithms such as network based proximity detection.