Indoor radio map maintenance by automatic annotation of crowdsourced Wi-Fi fingerprints

A radio map which consists of a number of Wi-Fi fingerprints has a tendency to became outdated over time. It is the result of both physical nature of the Wi-Fi signal and the variable indoor environment conditions which have an impact on received signal strength. Therefore, maintenance of the collected Wi-Fi signals data is essential for preserving quality of the Wi-Fi based positioning systems. However, usually it is the time and money consuming process. In the recent years number of solutions has been developed to address the problem. Unfortunately, in most of them human intervention is necessary. Our research concentrates on improving the quality of existing radio map by utilization of the crowdsourced Wi-Fi fingerprints that are post processed offline along with the logged pedestrian trajectory. The solution proposed in this paper focuses on the method which does not require human intervention, thus the end users do not have to report their locations. The results observed during the conducted research show the significant improvement of the positioning accuracy, in both surveyed and not surveyed locations.