Adaptive Autocorrelation Approach for Fingerprint-based Distance Dependent Positioning Algorithms in WLAN Indoor Areas

This paper addresses the adaptive autocorrelation approach for the fingerprinting-based distance dependent positioning algorithms (DDPAs) in wireless local area network (WLAN) indoor environment. As far as we know, although the DDPAs, like nearest neighbor (NN), K nearest neighbors (KNN) and weighted KNN (WKNN) algorithms, have been widely utilized for the indoor and outdoor location based services (LBS), the guarantee of location accuracy and precision has always been one of the significant compelling problems. Therefore, in response to this challenging task, the expected errors and associated confidence probabilities are mathematically deduced by the assumptions of logarithmic attenuation model and Gaussian distributions of received radio signal strength (RSS) at the receiver. However, because of the non line of sight (NLOS) property, time-varying interference and multipath effect, the measured radio strength varies a lot in the real-world indoor environment. Therefore, in order to fill this gap, a novel adaptive autocorrelation preprocessing approach is utilized to eliminate the singular strength from the original prestored radio map and improve the matching accuracy of DDPAs. Finally, compared with traditional DDPAs without adaptive autocorrelation preprocessing, the feasibility and effectiveness of the adaptive autocorrelation-based DDPAs are verified by approximately decreasing the average errors from 1.13m to 0.75m.