A High-Adaptability Indoor Localization Algorithm for Large-Scale BLE Sensors

Abstract

The rapid growth in demand for location-based services (LBS) is mainly due to the development of mobile devices and related services in recent years. As one of the LBS, the indoor positioning system (IPS) has become a research hotspot. Bluetooth technology has become one of the preferred indoor localization technologies due to its low cost and convenient installation. At present, most Bluetooth Low Energy (BLE)-based localization methods use fingerprinting approach. However, in the presence of large-scale BLE sensors, the traditional fingerprinting approach will take up a lot of computing resources or even can not be calculated. Meanwhile, the heterogeneity of mobile devices will also affect the localization accuracy. In this paper, we designed a dynamic selector to solve the problems caused by large-scale BLE sensors, and a dynamic weighted nearest neighbor (DWNN) indoor localization algorithm was proposed. Experiments show that the indoor localization method proposed in this paper designs a dynamic selector to solve the problem of large-scale BLE sensors. In addition, the device heterogeneity has been solved and higher localization accuracy has been achieved.