Outdoor Localization Framework with Telco Data

When Telecommunication (Telco) networks provide phone call and data services for mobile devices, measurement records (MRs) are generated to report connection states, e.g., signal strength, between mobile devices and nearby base stations. Telco outdoor localization is a technique to localize the mobile devices by using MR data. Unfortunately, city-scale Telco localization suffers from low localization accuracy, high cost of collecting sufficient MR samples, and noisy MR data. To tackle these issues, in this forum paper, we propose a machine learning-based Telco localization framework, consisting of three main components (localization models, the techniques to solve the data scarcity issue and to repair noisy data) and future directions.