Study in CSI Correction Localization Algorithm with DenseNet

Junna SHANG; Ziyang YAO

doi:10.1587/transcom.2021EBP3033

Open Access
Study in CSI Correction Localization Algorithm with DenseNet

Junna SHANG, Ziyang YAO

Full Text Views

120

Cite this

Free PDF (2.6MB)

Summary :

With the arrival of 5G and the popularity of smart devices, indoor localization technical feasibility has been verified, and its market demands is huge. The channel state information (CSI) extracted from Wi-Fi is physical layer information which is more fine-grained than the received signal strength indication (RSSI). This paper proposes a CSI correction localization algorithm using DenseNet, which is termed CorFi. This method first uses isolation forest to eliminate abnormal CSI, and then constructs a CSI amplitude fingerprint containing time, frequency and antenna pair information. In an offline stage, the densely connected convolutional networks (DenseNet) are trained to establish correspondence between CSI and spatial position, and generalized extended interpolation is applied to construct the interpolated fingerprint database. In an online stage, DenseNet is used for position estimation, and the interpolated fingerprint database and K-nearest neighbor (KNN) are combined to correct the position of the prediction results with low maximum probability. In an indoor corridor environment, the average localization error is 0.536m.

Publication: IEICE TRANSACTIONS on Communications Vol.E105-B No.1 pp.76-84

Publication Date: 2022/01/01

Publicized: 2021/06/23

Online ISSN: 1745-1345

DOI: 10.1587/transcom.2021EBP3033

Type of Manuscript: PAPER

Category: Navigation, Guidance and Control Systems

Authors

Junna SHANG
Hangzhou Dianzi University
Ziyang YAO
Hangzhou Dianzi University

Keyword

channel state information, indoor localization, fingerprinting, deep neural network, generalized extended interpolation, isolation forest

Cite this

Copy

Junna SHANG, Ziyang YAO, "Study in CSI Correction Localization Algorithm with DenseNet" in IEICE TRANSACTIONS on Communications, vol. E105-B, no. 1, pp. 76-84, January 2022, doi: 10.1587/transcom.2021EBP3033.
Abstract: With the arrival of 5G and the popularity of smart devices, indoor localization technical feasibility has been verified, and its market demands is huge. The channel state information (CSI) extracted from Wi-Fi is physical layer information which is more fine-grained than the received signal strength indication (RSSI). This paper proposes a CSI correction localization algorithm using DenseNet, which is termed CorFi. This method first uses isolation forest to eliminate abnormal CSI, and then constructs a CSI amplitude fingerprint containing time, frequency and antenna pair information. In an offline stage, the densely connected convolutional networks (DenseNet) are trained to establish correspondence between CSI and spatial position, and generalized extended interpolation is applied to construct the interpolated fingerprint database. In an online stage, DenseNet is used for position estimation, and the interpolated fingerprint database and K-nearest neighbor (KNN) are combined to correct the position of the prediction results with low maximum probability. In an indoor corridor environment, the average localization error is 0.536m.
URL: https://global.ieice.org/en_transactions/communications/10.1587/transcom.2021EBP3033/_p

Copy

@ARTICLE{e105-b_1_76,
author={Junna SHANG, Ziyang YAO, },
journal={IEICE TRANSACTIONS on Communications},
title={Study in CSI Correction Localization Algorithm with DenseNet},
year={2022},
volume={E105-B},
number={1},
pages={76-84},
abstract={With the arrival of 5G and the popularity of smart devices, indoor localization technical feasibility has been verified, and its market demands is huge. The channel state information (CSI) extracted from Wi-Fi is physical layer information which is more fine-grained than the received signal strength indication (RSSI). This paper proposes a CSI correction localization algorithm using DenseNet, which is termed CorFi. This method first uses isolation forest to eliminate abnormal CSI, and then constructs a CSI amplitude fingerprint containing time, frequency and antenna pair information. In an offline stage, the densely connected convolutional networks (DenseNet) are trained to establish correspondence between CSI and spatial position, and generalized extended interpolation is applied to construct the interpolated fingerprint database. In an online stage, DenseNet is used for position estimation, and the interpolated fingerprint database and K-nearest neighbor (KNN) are combined to correct the position of the prediction results with low maximum probability. In an indoor corridor environment, the average localization error is 0.536m.},
keywords={},
doi={10.1587/transcom.2021EBP3033},
ISSN={1745-1345},
month={January},}

Copy

TY - JOUR
TI - Study in CSI Correction Localization Algorithm with DenseNet
T2 - IEICE TRANSACTIONS on Communications
SP - 76
EP - 84
AU - Junna SHANG
AU - Ziyang YAO
PY - 2022
DO - 10.1587/transcom.2021EBP3033
JO - IEICE TRANSACTIONS on Communications
SN - 1745-1345
VL - E105-B
IS - 1
JA - IEICE TRANSACTIONS on Communications
Y1 - January 2022
AB - With the arrival of 5G and the popularity of smart devices, indoor localization technical feasibility has been verified, and its market demands is huge. The channel state information (CSI) extracted from Wi-Fi is physical layer information which is more fine-grained than the received signal strength indication (RSSI). This paper proposes a CSI correction localization algorithm using DenseNet, which is termed CorFi. This method first uses isolation forest to eliminate abnormal CSI, and then constructs a CSI amplitude fingerprint containing time, frequency and antenna pair information. In an offline stage, the densely connected convolutional networks (DenseNet) are trained to establish correspondence between CSI and spatial position, and generalized extended interpolation is applied to construct the interpolated fingerprint database. In an online stage, DenseNet is used for position estimation, and the interpolated fingerprint database and K-nearest neighbor (KNN) are combined to correct the position of the prediction results with low maximum probability. In an indoor corridor environment, the average localization error is 0.536m.
ER -