MENU

Advanced simulations for G.fast, Vectoring & Co

Advanced simulations for G.fast, Vectoring & Co

Technology News |
By Julien Happich



For several years, the cost-effectiveness of copper-based transmission systems has been the pillar of broadband internet access for domestic and corporate markets. Digital Subscriber Line (DSL), a family of technologies that provides broadband access over the telephone infrastructure, has continuously evolved aiming at providing faster and more stable services reusing the existing copper network. Its low investment requirements, great coverage and high data rates and stability make DSL a very attractive physical platform for supporting and offering services that demand high data rates.

The evolution of copper networks has been demarked by the necessity of counteracting two issues that limit their capacity: attenuation and crosstalk interference. Long loops, very common in former telephone networks to cover large territories, exhibit high attenuation of the transmitted signals, which reduces the Signal to Noise Ratio (SNR) at the receiver side and consequently, its capacity. This problem is aggravated by the usage of wide spectrum in new DSL systems that intends to add capacity in high frequency carriers to increase data rates; the attenuation levels are so high that high frequency carriers cannot carry any bit. On the other hand, the copper network is an end-to-end network that groups lines in an electromagnetically coupled medium: the cable-binder. Thus, services operating within the same cable-binder interfere with each other, mutually limiting their achievable data rates. This interference is known as crosstalk.

 

Vectoring mitigates crosstalk and obtains single line performance

DSL operators have transformed their copper network replacing long copper segments by optical fiber links. Thus, operators’ access system so called DSL Access Multiplexer (DSLAM) is installed at locations closer to the Customer Premise Equipment (CPE), reducing the signal attenuation and the number of mutually interfering services. Additionally, manufacturers have introduced techniques that allow DSLAMs and CPEs to compensate the crosstalk interference. This novel technique is known as Vectoring and uses Multiple-Input Multiple-Output signal processing to mitigate crosstalk and obtain single line performance. Therefore, transmission systems are not any longer seen as independent entities interfered by and interfering with coexisting counterparts. Instead, they are treated as collaborative entities that constructively use their mutual interaction, i.e. interference, to maximize their performance.


The transformation of the copper network and the new systems implementing advanced algorithms arise new challenges. The complexity of the network is higher since more DSLAMs are required to cover the same number of users, so the network topology becomes more diverse and heterogeneous. Moreover, regulative issues allowing multiple providers to serve users in the same cable-binder introduce a new degree of freedom with non-controlled services generating crosstalk interference that cannot be compensated by Vectoring. These factors may attempt against the roll-out success of new DSL technologies such as Vectoring and G.fast. Therefore, service and network operators demand tools that help them to accurately evaluate the potential of new copper-based technologies, while providing useful information for taking solid and strategic decisions for their business expansion and consolidation.

 

A customized simulation tool for DSL technologies

Through continuous research, Fraunhofer ESK has developed a customized simulation tool for Digital Subscriber Line (DSL) technologies that realistically model the complex network scenarios of the current copper access network. The tool takes the particularities of service provider networks (e.g. cable-binder type, loop length, use cases, etc.) into account and reproduces them using computer-based models of MIMO channels and signal processing units.

Fraunhofer ESK develops accurate MIMO channel models that reproduce access network scenarios and use them in extensive simulation studies to calculate the achievable data of real services under the given network conditions. Providers can then count on simulations results to take strategic decisions; for example, evaluate the performance improvements of Vectoring-VDSL2 in comparison with legacy technologies to upgrade their systems; calculate the achievable data rates of G.fast to estimate the possibility of offering new services that demand high data rates; or calculating the potential impact on and of competitors networks, and study possible strategies to reduce it.

Fig. 1: In the Access Network the DSLAM terminates the optical fiber, and the twisted copper pair establishes a connection to the wall plug unit. Finally the CPE provides access to the internet via wired Ethernet or Wireless LAN.
 

Figure 1 depicts a typical access network scenario for DSL services. As a consequence of progressive introduction of new technologies, broadband users at a building may be served by two different DSL technologies with access devices, i.e. DSLAMs, located at different distribution points. They may either belong to different providers or be part of the same network operator. In this example, we have chosen VDSL2 35b and G.fast since they are the most recent standardized DSL technologies. VDSL2 35b is an extension of the well-known VDSL2 that widens the operative spectrum up to 35 MHz. G.fast introduces new technological changes and enables to achieve gigabit data rates in short-range loops.


The present modelling and simulation approach allows us to obtain very accurate and detailed models of scenarios like the aforementioned. In addition, multiple aspects impacting systems performance can be considered, such as the number of users of each service,  the relative position between DSLAMs, the target data rates of services, the loop length of users and their position within the cable-binder, settings of transmission systems (transmitted power, spectrum, upstream/downstream ratio, etc.), among others. These aspects are varied and statistically analyzed to obtain average indicators and performance ranges that are important to define providers’ service coverage, portfolio and expansion strategies.

As an example of the usability of our simulation tool for such purposes, figure 2 shows the data rates that G.fast services with loop lengths between 30 m and 430 m can attain for different configuration settings of DSLAMs transmitters. A service provider can use these results to define the services that could be offered in accordance to the achievable data rates.

Fig. 2: The data rates which are achievable with G.fast depend on the loop length. These data rates can be optimized with different settings.

Additionally, if information of the loop length distribution of operator’s network were available, the percentage of potential clients achieving a given data rate, and consequently a given service, could be determined.

Another aspect that concerns broadband service providers nowadays is the coexistence of multiple operators within the same cable binders. As in figure 1, a service provider serving its users with VDSL2 35b from the cabinet may plan to upgrade its services to G.fast. The operator shall then replace a copper segment by optic fiber and locate its DSLAM in a distribution point closer to customer premises, so G.fast can provide higher data rates.


However, relocating DSLAMs impacts both former and upgraded services since topology changes increase interference power. Therefore, it is important to estimate the mutual impact that upgrading services may have to avoid degrading the performance of users that remain with the existing service and calculate the real improvements that deploying the new technology brings. Figure 3a and 3b shows the data rate achieved by 24 VDSL2 35b (red line) and G.fast systems deployed in the topology depicted in figure 1.

Fig. 3a/b: Figure 3a (above) and 3b (below) show the data rate achieved by 24 VDSL2 35b and G.fast systems deployed in the topology depicted in figure 1.

The results evaluate the case when 8 VDSL2 users are upgraded to G.fast (green line) and compare their performance with respect to the initial setup in which all services operate with VDSL2 35b. In addition, the performance of a very common technique used to mitigate the increment of interference known as Power Back-Off (PBO) is shown. Therefore, adding new systems without taking the existing system into account will result in a reduced performance of all systems, but simulation enables operators to evaluate the overall performance and get optimized settings for the existing and the new technologies.

This kind of analysis is not only valuable for estimating the benefits of deploying new technologies, but can also provide a means to optimize the configuration settings of existing systems. Such optimization processes and evaluation of the impact of new technologies for and with several European providers has been successfully carried out in ESK laboratories. This allows operators to take the right investment decisions.

 

About the authors:
Daniel Hincapie is Research Engineer at Fraunhofer Institute for Embedded Systems and Communication Technologies (ESK) – www.esk.fraunhofer.de

Mathias Leibiger is Group Manager Access- & In-house networks at Fraunhofer ESK

If you enjoyed this article, you will like the following ones: don't miss them by subscribing to :    eeNews on Google News

Share:

Linked Articles
10s