The future of broadcast video is the nDVR

In recent years, the most remarkable technological advances in the area of video delivery have been video-on-demand (VOD) and switched digital video (SDV). Today, there is much talk about the convergence of video, voice and data, which could provide ample benefits to content providers, service providers and consumers alike. But the benefits of convergence don’t stop there–other areas of convergence within the network can provide important benefits as well.

VOD has been around long enough to see some service evolution, but this evolution has been primarily limited to content delivered on-demand–at the consumer’s request. Today the evolution is expanding, and the next major step for VOD is broadcast content manipulation. Services like time-shifted programming, targeted digital ad insertion and ultimately true network digital video recording (nDVR) will help to redefine the consumer’s viewing experience, making it much more personal and flexible.

At the same time that VOD systems are ingesting more broadcast channels to support time-shifted services, MSOs are starting to deploy SDV networks in earnest. The next evolutionary step may involve SDV disappearing into the VOD network. Take an nDVR solution, subtract trick-modes, and add service group level resource management to allow multiple users to join the same session, and you have an SDV solution–within the VOD infrastructure.

More and more broadcasters are allowing some of their content to be time-shifted. There are systems deployed today with over 100 broadcast channels flowing through the VOD system to support time-shifted television. As more broadcasters realize the advantages of this offering, channels supported through the SDV network will also be available through the VOD system. Without convergence, routing the same broadcast channels through both the VOD system and SDV network will result in network redundancy without benefit.

Key improvements in VOD systems have made the concept of passing much of the digital channel line-up through the VOD system much more viable.

In many of the early VOD server solutions, content ingestion and streaming shared the same bus, memory and CPU resources, which typically resulted in a significant reduction of ingestion capacity as the active session count increased. Today improved VOD architectures apply dedicated resources to ingestion–separate from streaming–allowing the VOD system to ingest content at the maximum data rate independent of the number of active VOD sessions in use.

Improvements in hard disk speeds, intelligent use of RAM, faster computer I/O and high-performance interconnect technologies like Fibre Channel have all contributed to a reduction in content throughput latency. The latency of a broadcast flowing through the VOD server is on average five seconds, compared with the live broadcast. Improvements in this area will soon yield latency of less than one second.

VOD systems utilizing Commercial Off The Shelf (COTS) technologies are seeing greater rack densities in the areas of storage, streaming and ingestion. This allows the VOD system to support many new services without a significant increase in the number of racks required.

Convergent-evolution-Figure 1

These technological improvements make it possible to stream the entire digital channel line-up through the VOD system, but what are the business advantages? Removing network redundancies between VOD and SDV and building an infrastructure to support all manner of time-shifted television up to, and including true nDVR, support the business goals of improved network efficiency and ability to introduce new services.


Another advantage that could emerge from the convergence of SDV and VOD is in the area of ad insertion. Some of the long term goals of the advertising industry include placing targeted ads and conducting ad replacement in nDVR solutions. Unfortunately, the complexities of time-shifted services and SDV are causing cable operators to back away from the one ad targeted tool they have today–ad zones. VOD systems are now at the point where they are starting to manage ad inventory and manage ad splicing for VOD ad insertion. Adding these capabilities, along with broadcast channel ingestion and DPI-aware VOD servers, opens up new possibilities for the VOD system to restore the support of ad zones with an even finer resolution. This convergence of VOD, SDV and ad insertion allows targeted ads in both VOD and linear broadcast content.

An important architectural similarity between SDV and time-shifted solutions is in the area of MPEG conditioning. In both cases, the MPEG broadcast is clamped to a constant bit rate (CBR) output which improves the data switching through the network. Converting the broadcast to CBR eases the VOD server’s role in supporting broadcast ad insertion. Certain VOD servers already play a role in seamlessly transitioning between MPEG I-Frame based menus, previews and programs. It is not a technological leap for these same VOD servers to support the insertion of ads into a CBR MPEG broadcast stream.

VOD vendors that provide end-to-end systems are best positioned to support the convergence of SDV, VOD and ad insertion. These vendors have content management systems, databases and QAM session resource management software that can be adapted to support both linear and VOD ad insertion as well as SDV. VOD back office systems that employ Web Services can interface easily with ad management solutions utilizing emerging standards like the SCTE DVS 629, allowing the VOD system to provide targeted ads for both VOD content and linear broadcasts.

A converged VOD, SDV and ad insertion network may become a fundamental requirement to compete with IPTV networks. The main advantage of IPTV networks is the dedicated bandwidth per home and the ability to switch infinitely deep broadcast or VOD content into that dedicated bandwidth. The most comparable system an MSO has today that compares to a video delivery IPTV network is the VOD system. Like IPTV, the VOD system dedicates a specific amount of bandwidth to the home while in use, and switches various MPEG transport streams into that bandwidth.

As technological advancements make it possible for the VOD system to ingest the entire digital channel line-up, the main obstacle to a true nDVR is licensing. Because passing a broadcast through the VOD system does not represent any more of a legal issue than passing an MPEG transport stream broadcast through an MPEG multiplexer or Gigabit Ethernet switch, this is a good stepping-stone toward nDVR. By streaming the broadcast channels through the VOD/SDV system, the MSO will be able to build toward nDVR with each program that is licensed for time-shifted support. If the MSO can give broadcasters the tools to apply VOD trick-mode restrictions to nDVR content and to target and refresh linear ads, it is more likely that the broadcasters will more quickly embrace nDVR. These advantages, coupled with the improved content security and reduced consumer premises equipment costs associated with nDVR, make this solution very attractive to both the MSO and the broadcaster.

A converged VOD, SDV and ad insertion solution is a competitive advantage for the MSO and provides compelling incentives to the broadcasters to move more of their programming to the nDVR system.