Visible Light Indoor Positioning in a Noise-aware Environment

Abstract : Localization systems based on Visible Light Communication (VLC) are considered as good candidates for indoor environments, due to their high accuracy, low costs and the possibility of reusing existing infrastructures for both lighting and positioning. However, high level of environmental noises, mainly due to sunlight, significantly affect the performance of VLC positioning systems. A novel approach, for easily measuring environmental noises and compensating their effects, has been proposed in this work. Frequency Division Multiplexing (FDM) is adopted to divide the total bandwidth into a series of non-overlapping frequency sub-bands corresponding to each signal, while an estimation of Signal to Noise Ratio, obtained through real time Power Spectral Density measure, is exploited to compensate error positioning due to sunlight and other wide-band external optical nice sources. Proposed approach has been validated through experimental tests, carried out using a simple deployment of low power lamps, extremely low cost hardware and a Software Defined approach. In the region under test, receiver positions have been experimentally detected according to an improved accuracy in comparison with classical FDM approach, confirming the correctness of proposed technique, according to low Signal to Noise Ratio levels.
Complete list of metadatas

https://hal.archives-ouvertes.fr/hal-02022610
Contributor : Antonio Costanzo <>
Submitted on : Monday, February 18, 2019 - 10:01:04 AM
Last modification on : Thursday, February 21, 2019 - 4:53:46 PM
Long-term archiving on : Sunday, May 19, 2019 - 4:39:10 PM

File

Indoor_Positioning_WCNC (4).pd...
Files produced by the author(s)

Identifiers

  • HAL Id : hal-02022610, version 1

Collections

Citation

Antonio Costanzo, Valeria Loscri. Visible Light Indoor Positioning in a Noise-aware Environment. WCNC 2019 - IEEE Wireless Communications and Networking Conference, Apr 2019, Marrakesh, Morocco. ⟨hal-02022610⟩

Share

Metrics

Record views

55

Files downloads

89