Transfer Learning-Based Outdoor Position Recovery with Telco Data
This work addresses the challenge of accurate outdoor position recovery for mobile devices in telecommunications, offering a novel application of transfer learning to reduce data dependency and improve localization, though it is incremental in adapting existing transfer learning techniques to a specific domain.
The paper tackles the problem of outdoor mobile device localization using telecommunication data, which suffers from high data collection costs and poor accuracy in areas with scarce samples, by proposing a transfer learning framework called TLoc that transfers models from source to target domains, resulting in reductions of median errors by 27.58% to 49.22% compared to baseline methods on 2G GSM and 4G LTE datasets in Shanghai.
Telecommunication (Telco) outdoor position recovery aims to localize outdoor mobile devices by leveraging measurement report (MR) data. Unfortunately, Telco position recovery requires sufficient amount of MR samples across different areas and suffers from high data collection cost. For an area with scarce MR samples, it is hard to achieve good accuracy. In this paper, by leveraging the recently developed transfer learning techniques, we design a novel Telco position recovery framework, called TLoc, to transfer good models in the carefully selected source domains (those fine-grained small subareas) to a target one which originally suffers from poor localization accuracy. Specifically, TLoc introduces three dedicated components: 1) a new coordinate space to divide an area of interest into smaller domains, 2) a similarity measurement to select best source domains, and 3) an adaptation of an existing transfer learning approach. To the best of our knowledge, TLoc is the first framework that demonstrates the efficacy of applying transfer learning in the Telco outdoor position recovery. To exemplify, on the 2G GSM and 4G LTE MR datasets in Shanghai, TLoc outperforms a nontransfer approach by 27.58% and 26.12% less median errors, and further leads to 47.77% and 49.22% less median errors than a recent fingerprinting approach NBL.