Towards a practical indoor location matching system using 4G LTE PHY layer information

Predicting the location of a user in indoor settings in a practical and energy-efficient manner is (still) a very non-trivial task. The latest challenge in indoor localization is not to design specialized sensors but to design and implement practical data fusion methods using the already available technologies. Current state-of-the-art indoor localization techniques utilize Wi-Fi and a variety of sensors inside smart phones to predict user location. Some also require site-specific input such as indoor floor plans or the location of Wi-Fi access points. In this paper, we propose to use physical (PHY) layer information from 4G cellular network signals such as Reference Signal Received Power (RSRP) and Reference Signal Received Quality (RSRQ) to logically predict user location. Since the cellular signals are received by the smart phones at no additional cost, our methodology is very energy-efficient. We implement a prototype system in Android and evaluated it over 60 indoor locations. The prediction accuracy ranged up to 91% with an average localization error of less than 2.3m for any combination of 4G PHY layer information. The results show promise for improvements in current indoor localization systems using cellular signals.