ECE 6543 - Fiber Optic Networks
How to upgrade currently deployed TDM-based networks to next generation WDM-based access networks?
I - Challenges of next-generation optical access networks
Optical access has long been considered as the best solution to the problem of upgrading current congested accessnetworks to ones capable of delivering future broadband integrated services such as triple-play and HDTV. These types of services require bandwidths and QoS that are currently not available with traditional xDSL copper-based systems. Optical conversion of networks has already been performed in backbone networks, WANs, MANs, and even some LANs successfully. Thus, revamping access networks would be thelast step to the future all-optical network revolution. The problem we are raising is about how to move from the existing technology and structure to a new one which would be cost-effective, practically feasible and improve performance in order to tackle the upcoming subscriber’s demands.
Since the beginning of the 21th century, TDM-based Passive Optical Networks (TDM-PONs) have been deployed andare considered as an economically feasible solution for Fiber-To-The Home access (FTTH) . TDM-PONs work with one optical channel for multiple users and use Time Division Multiplexing to share the resources between them. This technology is therefore bandwidth limited. So, the next-generation of optical networks relies on integrating Wavelength Division Multiplexing (WDM) in the accessnetworks. This active field of research has introduced several architectures in recent years which show the technical feasibility of the WDM-PON . In these kinds of architectures, each ONU works on a dedicated wavelength and can therefore take advantage of the huge bandwidth provided by fiber-optic links. We will focus on one of these architectures: the SUCCESS-HPON. It is novel because it provides anefficient way to migrate smoothly from TDM-PONs to WDM-PONs .
Our approach for the rest of this paper is structured as follows: in section 2 we will introduce the general architecture of the SUCCESS-PON and focus on some novel elements introduced by its design that we consider as crucial, such as scheduling algorithms and continuous wave modulators. In Section 3, we will discuss theoptimization and upgrading of existing networks to WDM-PON-based architectures by analyzing the key components of these architectures: tunable lasers and arrayed waveguide gratings (AWG). Finally, we will conclude our synthesis by providing some elements of future research directions and technology evolutions.
II – The SUCCESS-HPON
SUCCESS–HPON stands for Stanford University aCCESS Hybrid(WDM/TDM) Passive Optical Network. We elected to study this architecture because, on one hand, it permits the reuse of existing network elements (TDM-PONs), and on other hand, it is highly cost-effective in that it uses bidirectional transmissions on the same wavelength and on the same fiber. Moreover, there is no need of expensive tunable transceivers at the ONUs since they send upstream data bymodulating continuous waves (CW) provided by the OLT (Central Office). The general architecture of the SUCCESS-HPON is shown on figure 1.
Furthermore, several improvements were made to the originally proposed scheduling algorithm that considerably enhance the measured performance of scheduling . Providing efficient scheduling is one of the key points of future WDM-based PONs because it isunthinkable to dedicate permanently one wavelength per user. Indeed, like in mobile networks, fiber network resources (number of channels, fiber-optic links, routing capabilities) are not unlimited and should therefore be handled intelligently. Thus, resources have to be assigned dynamically depending on each ONUs instant activity and needs. The innovation brought by the SUCCESS-HPON is to schedule...