Device-free and privacy preserving indoor positioning using infrared retro-reflection imaging

Santo, Hiroaki; Maekawa, Taira; Matsushita, Yasuyuki

doi:10.1109/percom.2017.7917860

Cited by 19 publications

(10 citation statements)

References 31 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Nowadays, there are a large number of applications for ILBS in market. And researchers have carried out a large number of studies on indoor positioning technologies and developed corresponding indoor positioning systems, such as the infrared [ 1 ], ultrasonic [ 2 ] or sound [ 3 ], radio frequency identity [ 4 , 5 ] (RFID), ZigBee [ 6 ], Wi-Fi [ 7 , 8 ], Bluetooth [ 9 ], microelectromechanical systems (MEMS) sensors [ 10 ], ultra-wideband [ 11 ] (UWB), geomagnetic field [ 12 , 13 ], visible light [ 14 ], computer vision [ 15 ], pseudolites [ 16 ] (PL), etc. However, due to the complexity of the indoor environment, it is usually difficult to provide a satisfactory level of accuracy in most applications.…”

Section: Introductionmentioning

confidence: 99%

An Adaptive Weighted KNN Positioning Method Based on Omnidirectional Fingerprint Database and Twice Affinity Propagation Clustering

Wang

et al. 2018

Sensors

View full text Add to dashboard Cite

The human body has a great influence on Wi-Fi signal power. A fixed K value leads to localization errors for the K-nearest neighbor (KNN) algorithm. To address these problems, we present an adaptive weighted KNN positioning method based on an omnidirectional fingerprint database (ODFD) and twice affinity propagation clustering. Firstly, an OFPD is proposed to alleviate body’s sheltering impact on signal, which includes position, orientation and the sequence of mean received signal strength (RSS) at each reference point (RP). Secondly, affinity propagation clustering (APC) algorithm is introduced on the offline stage based on the fusion of signal-domain distance and position-domain distance. Finally, adaptive weighted KNN algorithm based on APC is proposed for estimating user’s position during online stage. K initial RPs can be obtained by KNN, then they are clustered by APC algorithm based on their position-domain distances. The most probable sub-cluster is reserved by the comparison of RPs’ number and signal-domain distance between sub-cluster center and the online RSS readings. The weighted average coordinates in the remaining sub-cluster can be estimated. We have implemented the proposed method with the mean error of 2.2 m, the root mean square error of 1.5 m. Experimental results show that our proposed method outperforms traditional fingerprinting methods.

show abstract

Section: Introductionmentioning

confidence: 99%

An Adaptive Weighted KNN Positioning Method Based on Omnidirectional Fingerprint Database and Twice Affinity Propagation Clustering

Wang

et al. 2018

Sensors

View full text Add to dashboard Cite

show abstract

“…Arrival (TDoA), Received Signal Strength (RSS), and Channel State Information (CSI) [2]- [5]. The other category is to capture images with a video camera (color camera or infrared camera) and use image processing techniques to estimate the location of the objects [6], [7]. The radio-wave and the video-camera localization methods have their own advantages and disadvantages.…”

Section: Introductionmentioning

confidence: 99%

Deep Learning-Based Indoor Localization Using Received Signal Strength and Channel State Information

2019

View full text Add to dashboard Cite

Indoor localization has received wide attention recently due to the potential use of wide range of intelligent services. This paper presents a deep learning-based approach for indoor localization by utilizing transmission channel quality metrics, including received signal strength (RSS) and channel state information (CSI). We partition a rectangular room plane into two-dimensional blocks. Each block is regarded as a class, and we formulate the localization as a classification problem. Using RSS and CSI, we develop four deep neural networks implemented with multi-layer perceptron (MLP) and one-dimensional convolutional neural network (1D-CNN) to estimate the location of a subject in a room. The experimental results indicate that the 1D-CNN using CSI information achieves excellent localization performance with much lower network complexity. INDEX TERMS Indoor localization, deep learning, convolutional neural network (CNN), received signal strength (RSS), channel state information (CSI).

show abstract

“…[6][7][8] Many VLCIPSs use photodiodes as receivers of signals from light-emitting diodes (LEDs) and calculate positioning coordinates from the strengths of these signals. [9][10][11][12][13][14][15][16][17] Unfortunately, photodiodes are not frequently used and consumers cannot be required to use a photodiodes of receiver to configure store navigation devices. Since mobile devices are already considered to be personal necessities, in this work, the camera in a mobile device is used to receive VLC signals.…”

Section: Introductionmentioning

confidence: 99%

A high accuracy and efficiency indoor positioning system with visible light communication

Chen

2018

Trans Emerging Tel Tech

View full text Add to dashboard Cite

Positioning technology is important to the development of location‐based services. This work proposes a highly accurate and efficient indoor positioning system that uses visible light communication technology. The system uses a fingerprinting positioning mechanism that is based on a mobile device. A dynamic k‐NN (DKNN) mechanism is developed to improve upon k‐nearest neighbor (k‐NN) positioning, which is based on fingerprinting. This work also proposes an integrated angle of arrival (AOA) and fingerprinting mechanism for three‐dimensional positioning, which reduces the time that would otherwise be required for fingerprinting training and improves the accuracy of the indoor positioning system. When used for fixed‐height two‐dimensional positioning, the mean error distance of the DKNN fingerprinting positioning mechanism is approximately 0.15 cm. The AOA mean error distance is approximately 11.43 cm. The traditional k‐NN mean error distance is approximately 1.47 cm. Ignoring the height error, the integrated AOA and fingerprinting positioning mechanisms have a mean error distance of approximately 2.43 cm; the traditional AOA mean error distance is approximately 11.10 cm, and the fingerprinting mean error distance is approximately 2.74 cm. When the height error is considered, the AOA and fingerprinting fusion positioning mechanisms have a mean error distance of approximately 2.83 cm; the AOA mean error distance is approximately 11.62 cm, and the fingerprinting mean error distance is approximately 4.15 cm. The experimental results therefore show that the proposed mechanism provides the highest positioning accuracy.

show abstract

Device-free and privacy preserving indoor positioning using infrared retro-reflection imaging

Cited by 19 publications

References 31 publications

An Adaptive Weighted KNN Positioning Method Based on Omnidirectional Fingerprint Database and Twice Affinity Propagation Clustering

An Adaptive Weighted KNN Positioning Method Based on Omnidirectional Fingerprint Database and Twice Affinity Propagation Clustering

Deep Learning-Based Indoor Localization Using Received Signal Strength and Channel State Information

A high accuracy and efficiency indoor positioning system with visible light communication

Contact Info

Product

Resources

About