| Author: | Hai, On |
| Title: | Enhancement on accuracy of indoor visible light positioning by using k-nearest neighbours machine learning approach |
| Advisors: | Yu, Changyuan (EEE) |
| Degree: | M.Sc. |
| Year: | 2023 |
| Subject: | Optical communications Indoor positioning systems (Wireless localization) Machine learning Hong Kong Polytechnic University -- Dissertations |
| Department: | Department of Electrical and Electronic Engineering |
| Pages: | 65 pages : color illustrations |
| Language: | English |
| Abstract: | A novel technology of Li-Fi (IEEE 802.15.x) or Visible Light Communication (VLC) by means of visible light frequency spectrum as media of data transmission instead of using the conventional radio frequency band is one way of optical wireless communication (OWC). Well known that one of the promising advantage of the technology user friendly of enjoying the high speed of data transmission. Li-Fi data transmission rate have higher than ordinary 4G/5G over thousand times, ideally available to reach up to 1 Tbit per second transmission rate. Other robustness of VLC like clean energy, highly secured and free of interference noise, a value-added integration with the existed lighting equipment is widespread applicable at indoor restricted environment regarding high security of military, government and finance institutions, interference resisted features of airplanes and hospitals. Not only for indoor communication with a high transmission speed, VLC but derives a novel optical signaling technology application at indoor localization, which is Visible Light Positioning (VLP), it is an evolution of positioning using LED lamps. With variant advantages, VLP is established in inter-communications system widely used at indoor mobile robotic clusters system and indoor navigation system comfort our daily lives. Our practical work simulates an indoor VLC and VLP system, designs 4 LEDs suspended in ceiling as light transmitter source through line of sight (LOS) or directed channel (DC). To describe these used algorithms of CDMA as modulation of signal trans receiving, Fingerprint for Real Positions (RPs) original database in offline, Received Signal Strength (RSS) in Triangulation (i.e. Trilateration), as well as Signal-Noise-Ratio (SNR) analysis, finally by implementing k-nearest neighbours (k-NN) machine learning (ML) technique for enhancement of positioning performance. In accordance with the MATLAB simulation results, address the problem due to indoor under the impact of received power attenuation, at the corner or edge areas of low light intensity, due to signaling RSS or SNR depends on received power (or light intensity), dimmer environment of the concerned area increases the positioning error leading to plenty of unclear and scattered inaccurate positioning estimation. Main contribution to enhance the positioning accuracy of RSS-fingerprint, this thesis converges a solution by proposing k-NN ML to alleviate these shortcomings. While k-value as number of nearest neighbours critically affects the positioning performance, with MATLAB program to estimate the appropriate k-value in advance. The simulation illustrates by the result graphs that positioning performance has been improved after k-NN implement. The central part positioning is solidly firm almost the same well mated with the fingerprint original database. The positioning error is also improved in a large extent at dimmer area at the room corners or edges. This thesis consists of four chapters, firstly overview how the VLC and VLP optimize our daily lives. Next is to elaborate the details for VLC signal modulation and positioning technology related theories and algorithms (CDMA, RSS, fingerprint, k-NN). In third chapter, utilize these techniques through simulation works to introduce how to enhance VLP accuracy via k-NN ML approach. In the final chapter, share the VLC and VLP technologies potential orientations and applications toward the future innovative scientific research schemes and eco-system development. |
| Rights: | All rights reserved |
| Access: | restricted access |
Files in This Item:
| File | Description | Size | Format | |
|---|---|---|---|---|
| 8262.pdf | For All Users (off-campus access for PolyU Staff & Students only) | 2.53 MB | Adobe PDF | View/Open |
Copyright Undertaking
As a bona fide Library user, I declare that:
- I will abide by the rules and legal ordinances governing copyright regarding the use of the Database.
- I will use the Database for the purpose of my research or private study only and not for circulation or further reproduction or any other purpose.
- I agree to indemnify and hold the University harmless from and against any loss, damage, cost, liability or expenses arising from copyright infringement or unauthorized usage.
By downloading any item(s) listed above, you acknowledge that you have read and understood the copyright undertaking as stated above, and agree to be bound by all of its terms.
Please use this identifier to cite or link to this item:
https://theses.lib.polyu.edu.hk/handle/200/13740