基於數位微鏡設備之動態視界可見光通訊系統

Abstract

隨著LED以及通訊技術的發展，可見光通訊(VLC)在過去十年間有大幅的進展。近幾年的研究主要面向於增加可見光通訊的傳輸距離以及穩定度。然而，可見光接收器的設計卻是將可見光通訊實施於現實系統的一大問題。光電二極體是一個最常使用於可見光通訊系統的光接收器。但由於光電二極體會接受可視角內(FOV)所有光的特性，使得接收到的訊號容易受到環境光的干擾，也因為這個原因，傳統上的可見光接收器無法在同時有多個傳輸端的環境下工作，這是無法應用在實際無線通訊上的。因此，在論文中我們針對這兩個問題設計了一個動態調整可視範圍的可見光接收器，它能支持高速通訊並能區別多個傳輸端同時發送的訊號，甚至不需要利用複雜的多工技術就能達到多傳輸端通訊。我們利用壓縮感知(Compressive sensing)快速的偵測傳輸端的位置，接著透過數位微鏡設備(DMD)動態調整接收端的FOV來減少受到的干擾，並提出使用控制訊號在多傳輸端的情境下準確偵測所有傳輸端的位置。最終，實驗結果顯示，我們所設計的可見光接收器可以在太陽光干擾中相較於傳統接收器減少約10%解碼錯誤率、幾乎完全地消除了來自於其他傳輸端的干擾並且即便傳輸端在偵測中移動了2個像素的距離仍然可以偵測出傳輸端的位置。

As LED and communication technology advances, visible light communication (VLC) is improved by past research for decade. However, receiver is one of the main problems of VLC system to deploy in reality. Photodiode is the most commonly used receiving component of a VLC system. Due to characteristic of photodiode, photodiode-based receiver is vulnerable to many types of interferences. Furthermore, photodiode-based solution does not have the ability to separate and decode signals from multiple sources within the FOV, which is unacceptable for communication. In this thesis, we design and implement a dynamic-FOV VLC receiver which can support high-speed communications and have ability to separate signals from multiple transmitters broadcasting simultaneously. Our design further enables multi-transmitter communications without the need of a complex multiplexing technique. To mitigate interference, we leverage compressive sensing for fast detection of the transmitters, followed by dynamically adjustment of the size and the location of the receiver’s FOV. Moreover, we propose a method that exploits a control signal to accurately detect all transmitters under multi-transmitter scenarios. Finally, we adopt software solutions to synchronize received samples, instead of using hardware I/O, resulting in fast and accurate synchronization. Results of real-world experiments have shown that our system can reduce up to 10% decoding error rate under interference from sunlight, and also eliminate interference from other transmitter broadcasting at the same time. Our evaluation result also shows that the proposed system can still reliably detect the location of the transmitter even if the transmitters have movements up to 2 pixels during detection.