infrastructure solutions: The Triple Play Promise Becomes Reality

The potential of video on demand (VoD), real-time television or interactive TV has been discussed for many years. While the application and value have always been clear, the stumbling block has been the technology. It hasn't been fast enough or cheap enough to reach the mass market. However, due to significant progress on many fronts over the last few years, the desire for a "triple-play" service of POTS, high-speed Internet and video (both broadcast and on-demand) is being fulfilled by numerous telcos.

The value to triple-play service providers is clear: significantly lower customer churn and increased revenue per subscriber. In a recent study on cable companies that offer triple play, Morgan Stanley found subscriber churn was lower when the subscriber used two services. It was even lower when all three services were used in a single household. This reduction in churn resulted in annual cost savings of hundreds of dollars per subscriber.

Independent engineering consulting firms also have demonstrated the revenue potential with triple play (see table below). By increasing revenue nearly four-fold over single subscription voice-only customers, a carrier can add significant, profitable revenue.

With such an intriguing revenue model, telcos around the world are beginning to conduct a business and technical analysis of triple play. Carriers' first deployments of DSL have been of mixed profitability, especially in the United States where the telecom capital spending bubble created numerous competitors nearly overnight. The first business DSL service,SDSL, suffered due to the technology's potential to cannibalize profitable T1 revenue. The first residential DSL service, ADSL, struggled with provisioning and network management issues. Fortunately, ADSL now is being rolled out smoothly by most, if not all, carriers, and that experience will assist triple-play deployment -- creating a strong business case and rapid return on investment.

There are four key components of a video-over-DSL offering: the headend, the home set-top box, the access network and the video feeder network.

The headend (of which there usually is one per metro area or per regional fiber ring) captures the video content, encodes it for transmission to the home and integrates feature and administrative functions, including access control and billing. With today's standard encoders, broadcast quality or better channels can be compressed to operate in 3.5mbps to 4mbps data streams. HDTV today requires 9mbps to 12mbps.

The home set-top box decodes individual video streams for viewing on a television. The set-top box (usually via a remote control device) enables interaction between the viewer and the system. That includes everything from changing channels to Web surfing, interactive television guides, ordinary premium movies or videos. Older set-top boxes forced a carrier to always offer support for three TVs regardless of subscriber interest. Modern set-top boxes offer support for one, two or three televisions, and offer various in-home distribution approaches including coax and wireless.

The video access network must operate over a telco's existing copper-to-the-home infrastructure, and video signals must coexist with current POTS and broadband Internet services. Telcos are typically looking to offer two to three video channels per home, requiring 8mbps to 12mbps of bandwidth for simultaneous operation.

There are three types of DSL being considered for video deployment: ADSL, ADSL2 Plus (sometimes shortened to ADSL Plus) and VDSL. Selection between these technologies requires a trade-off between desired bandwidth, gauge of deployed copper wire and the distance between the xDSL equipment and the subscribers (sample rate/reach data for each type of technology is shown in the table above).

Using a recently available extension to the ADSL standard, known as S=1/2, on 24-gauge wire, ADSL is capable of driving 10mbps to nearly 10,000 feet. Today this is the DSL technology of choice for deployment given the state of the standard and the typical carrier serving areas. Many technologists say compression standards will continue to improve over the next few years, reducing broadcast content bandwidth to less than 1mbps. Proprietary approaches to compression are now achieving these kinds of results. With superior compression technology, telcos say ADSL S=1/2 will support future broadcast, HDTV and premium content video needs. For carriers interested in shortening their carrier serving areas to approximately 3,000 feet, VDSL offers enough bandwidth for HDTV. ADSL2 Plus, when standardized sometime in 2003, will offer a solution for carriers with carrier serving areas between 3,000 and 18,000 feet.

The video feeder network is the portion of the network connecting the headend, which outputs IP-over-gigabit Ethernet encoded content to the DSL equipment in the central office (or, more likely, remote cabinets with shorter carrier serving areas). In making a feeder network selection, the telco must first calculate bandwidth requirements and then map those requirements into potential technology options (the table Bandwidth Requirements, below, shows a typical scenario).

In the above scenario, more than 7gbps are required for the feeder network. Given existing technologies, this requires an OC192, four OC48s in parallel or eight bonded gigabit Ethernet connections (price comparisons for the various options are shown in the table above).

The video content enters the feeder network at IP video streams and can be delivered over either network type. However, when economics are considered, the gigabit Ethernet is the clear winner -- especially when paired with a sub 50ms restoration technology such as Ethernet protection switching, which provides SONET-like protection without the SONET cost or overhead.

Russ Sharer is vice president of marketing and business development for Occam Networks Inc. Contact him at rsharer@occamnetworks.com.

OC-48, OC-192 GigE Relative Pricing

DSL Rate/Reach Information, 24 Gauge Wire