Due to the physical nature of copper, delivering IPTV over DSL brings a number of challenges. It is clear that, over time, fiber architectures will become the end-game to deliver IPTV to mass markets, but in the meantime, whenever available, DSL remains the most cost-effective and fastest way to deploy IPTV.

The three main challenges linked to delivering IPTV over DSL, are bandwidth limitations, line stability, and power consumption.

The good news is that plenty of solutions exist to extend the reach of DSL networks, and to increase their ability to deal with very high-speed data and IPTV service requirements in an eco-efficient way. In this article, we will dive into the abovementioned issues and related solutions, and show that copper is indeed able to deliver on the promise of IPTV -- if the right deployment strategy and DSL innovations are being considered.

IPTV Uptake and Realities
Service providers around the globe have begun commercializing Triple Play to better compete with cable and satellite operators. While this is helping boost average revenue per user (ARPU), it is placing unprecedented stress on traditional high-speed Internet infrastructures.

Triple Play services comprise high-speed Internet access, voice, and video  delivered over a common architecture. The video component includes traditional broadcast services, video on demand and new, interactive, IP-based offerings. The content itself may be standard definition or ultra-vivid (and high-bandwidth) high definition. All of this falls into the category of IPTV, because it is delivered digitally over a converged IP network.

IPTV has been launched by major service providers worldwide, and its popularity with consumers is on the rise. The number of subscribers is growing exponentially from an estimated 10 million by the end of 2007 to 55 million in 2010. (See Figure 1.)

Figure 1. The number of subscribers is growing exponentially from an estimated 10 million by the end of 2007 to 55 million in 2010.

Obviously, delivering Triple Play services places a number of demands on the network. Video in particular has a significant impact:

• Bandwidth must increase to accom-modate personalization, interactivity and the proliferation of standard- and high-definition content. While basic Internet services and applications (such as web browsing and e-mail) generally have modest peak bandwidth requirements, the delivery of Internet peer-to-peer services - and especially video services - has dramatically increased bandwidth requirements. For instance, standard-definition TV requires approximately 3Mbps sustained bandwidth per stream; for high-definition content, demand increases to 8Mbps per stream. And what’s more, demands are expected to become even greater, as consumers acquire and add multiple high-definition television sets to their homes.1

• Reliability and stability must be guaranteed. After all, the delay or loss of video and voice packets may degrade the quality of these services to unacceptable levels. Customers will notice the errors. And the more errors they observe, the more complaints they’ll voice. This may compromise customer satisfaction and loyalty and increase help desk costs.
(See Figure 2.)

Figure 2. IPTV errors fuel customer complaints.

• Energy consumption needs to be reduced. In today’s economic environment, and linked to the current ecological mindset, energy consumption is a focal point of attention.

In the next sections, we will look at each of these requirements and potential solutions one by one.

Requirement 1:
Increased Bandwidth Support
As Triple Play becomes more user-centric, and as families start to simultaneously employ a richer mix of applications, the need for bandwidth grows. Various access network topologies exist to meet this demand, depending on an operator’s subscriber base and density, as well as the existing infrastructure.

Today’s “magic bandwidth number” for a successful IPTV service delivery is generally accepted to be in the range of 20 to 30 Mbps. This allows for 2 high-definition TV (HDTV) channels and a standard-definition TV (SDTV) channel, as well as voice and high-speed data traffic. Many operators envisage 50 Mbps further into the future.

The delivery of 20 Mbps or more to all subscribers generally requires a mix of copper and fiber in the access network. The exact mix, also called “fiber to the most economic point”, depends on the service blend, network topology, and legacy infrastructure. (See Figure 3.) In general, fiber-to-the-node (FTTN) and fiber-to-the-building (FTTB) combined with very high-speed digital subscriber line 2 (VDSL2) are more cost-effective in brownfield deployments, while fiber-to-the-home (FTTH) is more cost-effective in greenfield situations.

Figure 3. FTTx Fiber Migrated Models.

DSL technology is evolving continuously. Its latest flavor, VDSL2, adopts a number of improvements to better tackle the various bandwidth needs that result from IPTV service offers. It supports native Ethernet transport and provides tools for improved Triple Play stability. It also supports a wide range of deployment scenarios by using a profile concept. Each profile is linked to a specific bandwidth (from 8 MHz up to 30 MHz) and to different power and power spectral density (PSD) mask constraints. It is also optimized for a specific reach and rate. As such, VDSL2 pushes the maximum achievable bit rates above all other DSL flavors.

Today, commercially available VDSL2 solutions exist that deliver 50 Mbps sustained (100 Mbps peak) capacity over suitably short loops, using new, high-capacity line cards and supporting VDSL2 bonding. The bonding capability allows for increased bandwidth delivery at greater distances, enabling service providers to continue a rich service offering to all subscribers -- all while extending the life of their copper plant.

In Belgium, for instance, the “Broadway Project” of incumbent operator Belgacom builds on a state-of-the-art, FTTN VDSL2 access architecture that fully leverages Belgacom’s existing assets. The service provider has thus been able to cover (in less than 2 years time) more than 60% of the Belgian households with a sustained, video-grade throughput of 20 Mbps, providing its customers with a full array of Triple Play services including multiple IPTV streams, faster Internet connectivity, and HDTV.

Requirement 2:
Ensuring Line Stability and Quality of Service
As described in the requirements section above, basic Internet applications have modest, best-effort service needs that allow data packets to be discarded or delayed when congestion occurs (for instance, due to heavy traffic). The delay (or loss) of video and voice packets, however, may degrade the quality of these services unacceptably, causing pixelization and screen freeze.

It is clear that line stability has a huge impact on the end-user's quality of experience (QoE). A digital subscriber line made unstable by crosstalk, for example, can cause service degradation and even service interruption. Fortunately, continuous innovations allow service providers to significantly increase the capability of their copper networks to deal with very high-speed data and IPTV quality of service requirements.

Increasing the signal-to-noise ratio (i.e., using a higher noise margin) is the traditional approach to stabilizing lines that are impacted by crosstalk, but this typically results in a loss of bit rate, something that is not well suited in today's Triple Play world.

One solution, developed by Alcatel-Lucent, is called Smart DSL, targeting operators deploying higher-speed services over asymmetric DSL (ADSL) and very high-speed DSL (VDSL) lines. Basically, Smart DSL technology uses Artificial Noise (on ADSL) or Virtual Noise (on VDSL) to mask noise, and results in higher stable bandwidths for IPTV service delivery. This achieves optimal line stability with minimal bandwidth loss. It also allows operators to expand coverage of their copper networks for the deployment of IPTV and other premium bit rate services.

Future bandwidth and coverage improvements with DSL rely on the mitigation of crosstalk picked up by DSL pairs sharing the same binder. One promising technology is VDSL2 vectoring (also known as dynamic spectrum management level 3 [DSM3]). Vectoring is based on cross-talk cancellation within the binder, and could deliver bit rate gains of 25% to 100%. Because DSM requires some form of coordination among all users sharing a binder, these techniques are far from trivial. While DSM is a promising technology, line stability at high bit rates is one of bottleneck of DSM. Another non-technical issue such as loop unbundling, where different twisted pairs are operated by different companies, adds further complications to the DSM solution. These issues must be resolved before telcos can exploit the full potential of DSM.

Requirement 3:
More Bandwidth With Less Power
Given today's economic environment, and the public's current ecological mindset, energy consumption is a substantive issue for providers. Lowering the power consumption of IPTV not only helps operators reach their eco-sustainability targets, but also reduces their operating cost.

Savings can be made by using the latest generation of DSL line cards, which can consume up to 20% less power than the previous generation, all while supporting higher densities/capacities (i.e., more users, and higher bandwidths per line card).

From an outside plant perspective, a distributed DSLAM concept and the use of custom-made cabinets can make quite a difference as well. Concretely, a distributed DSLAM architecture consists of remote DSLAMs that are managed as remote line cards, connected to a central host. This model not only results in reduced aggregation capital expenditure and in cost-effective operations, but also lowers the overall power consumption since the remote nodes share a central controller.

Today's standards that allow modems to reduce power consumption help the cause. Modems can reduce power either when user traffic is low (L2 mode) or when the service rate can be achieved with excess margin (maxSNRM). Whereas these varying modem signal levels result in varying crosstalk noise that can destabilize the network, these issues can now be overcome using Smart DSL technology, enabling a significant reduction (-25%) of the modems' power consumption across a service provider's subscriber base.

It's All About Problem-Solving
Offering Triple Play services over a DSL network poses radically new challenges compared to offering simply high-speed Internet. Issues of bandwidth, multiple application support, quality of service/experience, and ecological footprint must all be considered.

Thanks to a number of recent innovations, DSL networks are indeed perfectly capable to support end-users' needs for high-speed broadband services, for a number of years to come.

Endnote
1. According to Nielsen Media Research [www.nielsenmedia.com, "Nielsen Reports Television Tuning Remains at Record Levels", New York, October 17, 2007], video services will be used more frequently than e-mail or the web in coming years. Currently, the average United States household has 2.8 televisions (though only 2.5 occupants, on average) and occupants spend 8 hours and 14 minutes per day watching.

 About the Author
Stefaan Vanhastel is Product Marketing Manager FTTx for Alcatel-Lucent. He is responsible for the global marketing of Alcatel-Lucent DSL products and technology. He was appointed to his current role in 2005. Prior to joining Alcatel-Lucent, Stefaan worked at a start-up company specialized in network and datacenter infrastructure management, and in the Broadband Communications group at the University of Ghent (Belgium). He has managed research projects focusing on network performance and network management. For more information, visit www.alcatel-lucent.com or email Stefaan.Vanhastel@alcatel-lucent.be.

What is your experience with this? Tell your fellow readers now!