Achieving Cost-Efficient Indoor Fingerprint Localization on WLAN Platform: A Hypothetical Test Approach

Received signal strength (RSS) is a typical type of measurements used for indoor fingerprint localization on wireless local area network platform. To make good use of RSS information, we rely on the hypothetical test approach to perform localization with the optimized access points (APs). Specifically, in offline phase, the operating characteristics function is used to minimize the sample capacity of fingerprints at each reference point, and meanwhile the APs are optimally selected based on the concept of information gain criterion. Then, in online phase, the F-test and T-test approaches are used to conduct the RSS variance and mean test, respectively, with the purpose of achieving RPs matching, namely coarse localization. After that, the density-based spatial clustering of applications with noise is developed to realize fine localization with the improved accuracy performance. The extensive experimental results demonstrate that the proposed system is able to avoid the blindness of fingerprints collection as well as improve the effectiveness of fingerprints matching especially under the small sample capacity of fingerprints.