When a mobile device connects to a telecom network (e.g., GSM 2G, CDMA 3G and LTE 4G), the network generates measurement report (MR) data containing signal conditions to support communication services. Using MR data to predict the accurate position of a mobile device is an important problem in Telco industry, called Telco localization problem. Although the literatures have proposed various MR-based localization algorithms, it is unclear how such algorithms perform in terms of localization precision. We develop a tool named STLT, which supports comprehensive performance study of a broad category of the state-of-art localization algorithms. STLT provides flexible training/testing data division, deep parameter turning as well as three visualization modes to display localization results. Through the demonstration of STLT on three real MR datasets provided by one of the largest Telco operators in China, we have three interesting findings that could inspire and enhance future research in Telco location.
[1]
Tao Gu,et al.
City-Scale Localization with Telco Big Data
,
2016,
CIKM.
[2]
Mohamed Ibrahim,et al.
CellSense: An Accurate Energy-Efficient GSM Positioning System
,
2011,
IEEE Transactions on Vehicular Technology.
[3]
Erik G. Ström,et al.
RSS-based sensor localization with unknown transmit power
,
2011,
2011 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP).
[4]
Frank Klawonn,et al.
Multi-Layer Perceptrons
,
2013
.
[5]
Ramesh Govindan,et al.
Energy-efficient positioning for smartphones using Cell-ID sequence matching
,
2011,
MobiSys '11.