Moving away from all-satellite for content aggregation and distribution.

TiVo and the Internet have changed our media viewing habits and how we want to be entertained. Consumers are demanding choice and convenience, and media and entertainment executives have subsequently adopted an “any content, anywhere, anytime and on any device” approach. Achieving this vision – ensuring content is at the right place at the right time in the right format, and at the right cost – is critical to making these new business models viable.

New service offerings like Comcast’s project Infinity, which aims to extend its video library from 1,300 to more than 6,000 available on-demand movies a month by 2009 – each with different tiers of pricing; or Netflix, which is looking to make its library of 100,000 DVD titles available for streaming on a dedicated set-top box, necessitate advances in the acquisition, storage, access and delivery of very large, and increasingly high-definition (HD) media files. And with the daily introduction of new feature-length movies, television content, news, sports, Webisodes and user-generated content, the amount of content that needs to be aggregated and distributed for on-demand viewing is only increasing.

Currently, cable service providers rely on satellite to acquire video from content owners, including broadcast networks and movie studios for video-on-demand (VOD) services – a costly, labor-intensive approach that requires the use of multiple satellites and transponders.

Figure 1: IPTV – Content ingest for on-demand service.
Figure 1: IPTV – Content ingest for on-demand service.

With some budgets for satellite exceeding $700M annually, and the need to distribute more content with higher resolution, companies are looking to utilize their terrestrial (Ethernet/IP) network investment for several reasons:

• Weekly cable programming must be distributed to multiple countries and outlets in many formats to meet deadlines. Instead of using tape or satellite, broadcasters are turning to digital file-based workflows to transfer files from U.S. broadcast facilities to national and international drop points using their terrestrial infrastructure.

• As content becomes available from a wider array of partners, including VOD aggregators and aggregator services; Internet portals; gaming and social networks; and advertising, trailers and promotions from advertising agencies, these service offerings must be distributed to a wider array of channels, including broadband, mobile and portal partners.

• The economic advantages of switching to Ethernet delivery are significant. Distributing the “long tail” of VOD content becomes more and more costly, and distributing HD content over satellite is just not going to be effective, especially as more live broadcasts move from standard definition (SD) to HD and transponder time is needed to accommodate real-time content delivery.

Content, historically acquired via satellite as noted above, is stored in a server farm at the cable master headend – the place where live and on-demand content is aggregated from the various content owners. Once the on-demand content is aggregated at the master headend, it must be distributed to the regional headends and hubs to be closer to the subscribers.

Distributing the content to many remote servers for delivery to the subscriber follows the principles of content delivery networks (CDN). The distribution of content ensures that the network can support the subscriber base with a good quality of experience (QoE). When a subscriber requests a movie, it’s streamed to them over the cable network to their home, processed by the digital cable set-top box connected to their TV set, and then presented on the screen.

With the declining price of content – some of which can be acquired for as low as 99 cents to $1.99 on iTunes or Amazon; $1.99 from cable; or the “all-you-can-eat” models of TiVo, YouTube and Netflix – it becomes critical that costs are controlled to ensure delivery of the content does not exceed the cost of acquisition.

Figure 2: Cable network – Regional programming ingest.
Figure 2: Cable network – Regional programming ingest.

While popular new releases like “There Will Be Blood,” “Rambo” and “National Treasure 2” will be widely viewed, the costs of acquisition and distribution are amortized over many views.

Individual long tail titles by definition are accessed infrequently, so the costs to acquire, distribute and deliver these files need to be monitored closely to ensure that profit margins are maintained.

Watching these costs is especially important because even though individual “long tail” titles – “Citizen Kane,” perhaps, or documentaries – may not be viewed as frequently, in aggregate, long tail content can represent as much as 80 percent of the content requested by subscribers. It is simply not practical to populate HD versions of all of the long tail content using satellite bandwidth, however.

Swayed by cost savings and operational benefits, several U.S. cable service providers are working to move content from satellite-based delivery to terrestrial, including one cable operator that is targeting 700 remote headends, and a major cable operator in Texas that has built out a high-speed IP fiber network to distribute content from its master headend in Texas to all of its remote headends.

Despite the fact that many of the largest communications providers are consolidating operations into so-called superheadends, there is still a requirement to distribute VOD content closer to the home via a CDN to give customers a better QoE. While long tail content can be positioned closer to the superheadend (and probably not in it), popular content needs to be positioned closer to the home in a regional headend or possibly even the network hub.

Figure 3: Cable network – Ad insertion.
Figure 3: Cable network – Ad insertion.

As cable providers follow IPTV services’ lead in building out IP networks, they are increasingly able to take advantage of terrestrial digital video delivery, eliminating the cost of satellite and even taking advantage of shared network bandwidth.

But these networks – fast as they may be – still suffer from latency, mismanagement and downtime when moving on-demand files during peak viewing hours. Understanding the global view of a content distribution network – what files need to be aggregated and distributed and when – is critical to meeting service level agreements and ultimately monetizing content.

Digital media distribution management technology can help cable, Internet and IPTV service providers to accelerate, manage, automate and secure the movement of digital media, ensuring content gets to the right place, at the right time, in the most cost-efficient method possible.

To aggregate or ingest content in an IP network from content providers, and deliver it to a master headend, cable operators require an Internet connection with enough bandwidth engineered into the infrastructure to capture the content they need to ingest. This can be calculated by assessing how many gigabytes of content need to be moved, network capacity and the time that content needs to be there.

To distribute this content from the master headend to the remote headend, cable companies need an IP network. This can be proprietary, or leased from a service provider that will carry the aggregate content from source to destination within the time period that is required.

When making the move to an IP-based solution to support large files and using standard protocols (TCP, HTTP, FTP), it is important to realize that any latency and network congestion will dramatically lower bandwidth efficiency. For example, the effective bandwidth for a 100 Mbps link could be less than five percent over extended distances.

There are numerous wide area network (WAN) acceleration products on the market that can alleviate latency and congestion problems to enable large files to move through networks without assistance, available both on-demand and as an in-house solution.