從分時多工被動光網路(TDM-PON)升級到分波多工被動光網路(WDM-PON)的研究和發展

Abstract

由於頻寬的需求，被動光網路(passive optical network, PON)已經在世界各地的多個國家被使用。和一 般以銅線為基礎的接取網路比較，分時多工式的被動式光網路(TDM-PON)可以供應使用者更多的使 用頻寬。然而儘管TDM-PON在成本上比較便宜，它卻具有要和別人共享頻寬的缺點。為此，分波 多工式的被動式光網路(WDM-PON)被提出來以更加提升可用頻寬。在WDM-PON中，每個光網路 單元(optical network unit, ONU)都各自享有一組分別用來做為上傳信號和下傳信號載子的波長。因 此，在不需要和別人共享頻寬的情形下，每個光網路單元都能獲得更高的傳輸頻寬。由於WDM-PON 是未來極有發展潛力的技術，許多仍以銅線做為網路接取方法的地區都考慮直接將它們現有的網路 架構直接改成WDM-PON。不過對於TDM-PON已經大量發展的地區，這種全面性的改變相當困難。 這個原因在於以往投資的光網路架構必須被保護。因此，在不改變現有TDM-PON的前提下，一種 可以將TDM-PON架構漸進式地改成WDM-PON架構，或是讓兩種光網路共存的技術就變得很重要 了。

在這個計晝中，我們的重點有二個：（1)透過使用差分相位調變技術(differential phase shift keying, DPSK)和光濾波器，我們提出一種能夠調和TDM-PON和WDM-PON的方法，並實際透過實 驗展示它的可行性。我們架構中使用的光濾波器可以被預先裝在光網路單元以選取特定的下傳波長 並解調下傳信號。（2)我們透過再調變技術(remodulation)將下傳DPSK的信號重新做為上傳信號使 用。在這種技術中，波長的參考和控制由局端(central office, CO)提供。因此，在成本敏感的ONU中 並不需要使用特定波長或可調波長鐳射。但這種做法會讓上、下傳信號在使用相同波長並同時在一 根光纖上傳輸時受到較嚴重的雷利背向散射(Rayleigh backscattering, RB)而影響信號品質。於是，我 們進一步提出並展示結合波長漂移(wavelength-shift, WS)和再調變的技術，來減少雷利背向散射造成 波長重疊的雜訊。

這個計晝可以幫我們的夥伴建立一個由TDM-PON架構漸進式地改成WDM-PON的方法。我們 在這個計晝中的夥伴是 OCP Asia, Inc。OCP Asia, Inc 是 OPLINK Communications, Inc, USA 的成員公 司，OPLINK總公司位於美國加州，為一家在美國那斯達克(NASDAQ)掛牌的上市公司，其據點分 布於歐洲、美國、中國、以色列、加拿大、日本及台灣各地，在光通訊主被動元件產業的發展上， 具有相當高的知名度，並且是全球最大的光器件及模組供應商之一。公司致力於設計、生產、銷售 高性能光纖網路器件及集成光學模組，應用於光傳輸、光交換、光放大、城域網、寬帶接入、光纖 局域網等高速光通訊領域。這個計晝可以促進學述界和業界的合作，並讓想法和技術在研究者和工 程師間交流。

Passive optical network (PON) has already been deployed in many different countries all over the world to meet the huge bandwidth demand of broadband services. When compared with the copper-based access networks, these time-division-multiplexed (TDM) based PONs can offer high bit-rate to end-users. Although these TDM-PONs are cost effective, they suffer from the bandwidth sharing nature. As a result, wavelength-division-multiplexed (WDM) PON has been proposed to further increase the bit rate for users. In the WDM-PON, a dedicated pair of wavelengths is assigned to each optical networking unit (ONU), hence high bit rate transmission in both upstream and downstream directions can be easily guaranteed for each ONU. As the WDM-PON could be a future-proof solution, many places which are using copper-based access solutions are considered to deploy the WDM-PON directly. However, it may be difficult for some places where TDM-PONs have already been built. The past investments of these fiber infrastructures should be protected. Hence solutions for smooth mitigating from TDM-PON to WDM-PON without the need of changing the existing fiber infrastructure; or allowing the co-existence of TDM-PON and WDM-PON are highly desired.

In this project, we will focus on two areas: (1) we propose and demonstrate a migration scheme from TDM-PON to WDM-PON using differential phase shift keying (DPSK) with optical filtering. The optical filter can be pre-installed at the ONU to select the desirable downstream wavelength and to demodulate the downstream DPSK signal simultaneously. (2) Then, a signal remodulation scheme is used in which the downstream DPSK signal can be reused to produce the upstream signal. In this approach, the wavelength referencing and control is provided by the central office (CO), hence wavelength specific or wavelength tuneable sources are not required in the cost-sensitive ONU. However both the downstream and upstream signals have the same wavelength and share the same fiber, hence Rayleigh backscattering (RB) will severely affect the upstream signal. Hence we further propose and demonstrate a wavelength-shifting (WS) signal remodulation ONU to reduce the spectral overlap of the upstream signal with the RB noise.

This project can build up migration solutions from TDM-PON to WDM-PON for our partner in this project: OCP Asia, Inc, (a member company of OPLINK Communications, Inc, USA). The headquarter of OPLINK is in the United States and it is also a NASDAQ listed company. It has sales and R&D offices in Europe, the United States, China, Israel, Canada, Japan and Taiwan. It is one of the leading providers of optical networking components, modules and subsystems in the world. This project can also promote the collaboration between academic and industries, allowing idea and technologies transfer between researchers and engineers.