An Enhanced Location Estimation Approach Based on Fingerprinting Technique

Abstract

This paper aims to locate the target position by using RFID Readers and tags with minimum error. After investigating the existing fingerprinting location technique, we find that the distribution of RSSI-value at each sample point is fluctuant even in the same position. Therefore, we propose an enhanced algorithm to the existing deterministic algorithms. The proposed enhancement approach collects ranges of possible RSSI-values from either sample point or target object to build the probability model, which is different from original algorithm that only saves one single RSSI-value at each point into a database, then several sets of RSSI-values are got for further location by using permutation and combination. Unlike the existing techniques, more than one locations are computed and the final location is the weighted average of those computed locations. simulation results and experimental comparisons are shown, in order to prove the accuracy of location estimation is improved.