CC-DTW: An Accurate Indoor Fingerprinting Localization Using Calibrated Channel State Information and Modified Dynamic Time Warping

Abstract

Indoor wireless local area network (WLAN) based positioning technologies have boomed recently because of the huge demands of indoor location-based services (ILBS) and the wide deployment of commercial Wi-Fi devices. Channel state information (CSI) extracted from Wi-Fi signals could be calibrated and utilized as a fine-grained positioning feature for indoor fingerprinting localization. One of the main factors that would restrict the positioning accuracy of fingerprinting systems is the spatial resolution of fingerprints (SRF). This paper mainly focuses on the improvement of SRF for indoor CSI-based positioning and a calibrated CSI feature (CCF) with high SRF is established based on the preprocess of both measured amplitude and phase. In addition, a similarity calculation metric for the proposed CCF is designed based on modified dynamic time warping (MDTW). An indoor fingerprinting method based on CCF and MDTW, named CC-DTW, is then proposed to improve the positioning accuracy in indoors. Experiments are conducted in two indoor office testbeds, and the performances of the proposed CC-DTW, one time-reversal (TR) based approach and one Euclidean distance (ED) based approach are evaluated and discussed. The results show that the SRF of CC-DTW outperforms the TR-based one and the ED-based one in both two testbeds in terms of the receiver operating characteristic (ROC) curve metric, and the area under curve (AUC) metric.

LC-DNN: Local Connection Based Deep Neural Network for Indoor Localization With CSI

“…Full Connection 60 50 40 30 2 3 88 60 50 30 2 5 258 200 100 50 2 10 405 300 200 100 2 20 710 500 300 100 2 30 915 800 500 300 2 to our previous research, the time stability of CSI data is affected by the restart of transmitter and the change of environment [35]. Thus, in this paper, both of the measurements are conducted without restart of transmitters to keep the stable environment.…”

“…It provides effective information of position due to the stability over time and specificity over positions. Second, different antennas have different CSI features [35], [36]. Fig.1 shows CSI amplitudes on three antennas from 500 received packets at a single position, 30 subcarriers for each antenna.…”

Decision Tree Pattern Recognition Model for Radio Frequency Interference Suppression in NQR Experiments

“…The tree starts at a root node, where the data is divided into two smaller sets based on a binary “feature” test of the data—that is, upon whether some quantitative measure derived from the data is, for instance, greater or less than a threshold value. Features can be temporal- or frequency-based, or based on some characteristic of the probability distribution of the data [20,21,22,23]. Further successive binary splittings of the data are made, based on further feature decisions.…”