Characteristics of a large-scale WiFi radiomap and their implications in indoor localization

Due to the wide-spread of mobile devices such as smartphones, localization at both indoors and outdoors constitute a critical component in numerous location-based services. This paper considers indoor localization using WiFi fingerprints (FPs), which is a set of received signal strength (RSS) values from observable access points (APs). Typically, FPs are collected at several locations in a venue to create a WiFi radiomap, which is then used to estimate the location of a client device. However, most of previous studies evaluated their indoor localization algorithms and mechanisms on small-scale venues such as academic buildings, leaving us an important and practical question whether they perform well in large-scale environments with hundreds or thousands of APs. We surveyed a large-scale venue of Gangnam Coex underground mall in downtown Seoul, Korea, which might be sufficient to make most of existing solutions to suffer. This paper characterizes the radiomap of Gangnam Coex while highlighting the differences with that of a typical, small-scale academic building to facilitate the design of indoor localization algorithms for large-scale venues.