Improving accuracy in indoor localization system using fingerprinting technique

Indoor localization system is an important topic in the wireless navigation system. Many technologies, such as Wi-Fi, Bluetooth, and ZigBee, have been developed for the indoor localization system; however, these systems still demonstrate poor accuracy and performance. The indoor localization system using Wi-Fi signals is the best option for indoor localization because most of the buildings are covered by Wi-Fi access points. In this paper, we proposed a clustering algorithm to improve the system accuracy for indoor environment and reduce computational time. Fingerprinting technique was used to evaluate the performance of the proposed algorithm. Results show that the clustering algorithm achieves an average distance error of 2.4 m compared with the non-clustering algorithm, which has an average distance error of 3.4 m. In particular, the clustering algorithm improves the system accuracy by 41% compared with the conventional algorithm. Moreover, the clustering algorithm reduces the computational time of the system that requires computing the prediction of mobile user location.