Forget compression

While sports broadcasting is driving the HD trend in television broadcast technology, it is also highlighting underlying problems with the mass market transport of real-time and HD-quality rich media content. To provide the “it's like I was there” experience promised by HDTV, video must be delivered with the greatest and highest possible quality (e.g. no latency, jitter). The entire entertainment content and delivery value chain faces the ever-increasing consumer demand for uncompromising high quality. Video on demand, interactive gaming and new end-user TV appliances, such as mobile devices, are just some examples of where HDTV-quality expectations are driving the need for an agile, video-centric network infrastructure.

HDTV delivery to a mass market today represents an unprecedented opportunity for innovation since the dawn of the Internet. Equipment vendors are clamoring for technological advances that allow the production and contribution, distribution, and delivery of rich content with quality that meets user expectations.

In the production and contribution world, dedicated media networks connecting remote sites, digital production units and studios provide seamless video delivery in a lossless environment from the field to the studio. Broadcast networks are now beginning to demand a similar QoS capability from telecom providers. This is a must as rapidly improving TV quality has set user expectations at a level beyond even the smallest compromise. Users are demanding studio quality in the home.

Distribution issues

Content distribution provides an added challenge. Compressing content for bandwidth efficiency ultimately results in degradation. Transporting uncompressed HD content, however, is rare. Telecom operators have raced to install more fiber and hardware solutions, but many networks are still unable to support uncompressed video. As a result, the packet switching fabric is being stressed to the limit. These networks are designed for best-effort services and do not guarantee QoS. HDTV just compounds the problem.

Content delivery begs the question of whether there is enough delivery bandwidth to allow each end user to have the “like I was there” rich media experience of HDTV. Today's network infrastructures do not support the demand for video transfer as they did for data and voice, leaving everyone to compromise on quality. As a result, the promised HDTV experience is hardly a reality and won't be anytime soon. (See Figure 1.)

According to Informa Telecoms & Media, the number of homes worldwide with HD-capable televisions will jump from 48 million at the end of 2006 to 151 million by 2011.

Sports broadcasters are driving demand for HD content delivery. Research shows that consumer education is also improving. Both developments have led distributors to broaden their HD offerings and cable networks to accelerate their plans to jump into HD. This increase in demand has left network operators scrambling to pull even more fiber and upgrade networks to handle the influx.

The challenge

Many broadcasters are upgrading existing infrastructure so video can be transported over IP networks. IP is a best-effort packetized transmission protocol designed to make transport facilities efficient. The presumption is that if data is grouped into packets, with each packet having sufficient data in its overhead to identify its destination, then the packet will be free to travel across the network in an efficient manner determined by the intelligence of the switching transport network.

While IP has enabled expansion of video, audio and voice services, its packet-based protocol is generally not well suited for video. Content providers, rich media distributors and delivery networks must address a variety of performance issues that are inherent in IP networks and can cause a reduction in quality and service. Frequent issues include:

  • IP allows jitter, especially across large networks.
  • Mixed interface planes could create inefficient signal processing.
  • Live video needs a synchronous infrastructure.
  • Best-effort networks may result in packet collisions or lost packets.
  • Multicasting will eat up any available bandwidth.

In the world of video — particularly live video — dropped packets wreak havoc on transmissions. Lost packets during transmission could cause viewers to miss the winning goal in the World Cup or Super Bowl.

Reserving network switch resources is one way to ensure all video data packets arrive properly and the viewing experience remains intact. Various studies have looked at the connection between the size of an IP network and the quality of video transmission across the network. In order to control video quality across a growing IP network, these studies have found that portions of the network's capacity need to remain empty to prevent packet collision or congestion, which limits the viewing experience.

As the network grows, the relationship between the number of hops and the amount of network bandwidth left empty has a strong, negative correlation. Sending a video signal across one network hop results in the video signal being received with 100 percent quality. No network resources need to be reserved to counteract the adverse effects of packet collision. However, for a video signal traversing 16 hops, almost 90 percent of the network capacity must be reserved in order to maintain quality.

Is it possible to build a viable converged IP network capable of supporting broadcast-quality video? Because live video demands the highest priority in the network, the transmission medium should be one where other traffic will not interfere with or impede the video's transmission across the network infrastructure. As networks experience growing multicast group transport, the growth of the demand should not negatively affect the video quality.

IP, though not optimal for video applications, remains the most prevalent protocol for next-generation converged networking platforms. IP offers tremendous possibilities for network efficiencies. Meanwhile, Ethernet is driving down possible return on investment, which companies must rely on to maintain their position in the market. As a result, if IP is supplemented with next-generation algorithms and technologies, IP may find a fit in the rich media broadcast applications where live video is an important component for a network operator's product suite.

The opportunity

Recognizing that the inefficiencies of IP networks can negatively affect the viewing experience and negate the benefits of HD broadcasting, sports broadcasters are turning to a hybrid technology, Dynamic Synchronous Transfer Mode (DTM). DTM allows broadcasters to produce, distribute and deliver uncompressed HDTV to consumers without jitter, delay or reduced quality. DTM combines the best of circuit switching (required for transporting large, uncompressed video signals) and the best of packet-switching, delivering full-featured and economical HD experience. (See Figure 2.)

Designed to improve some of the shortcomings inherent in IP networks, DTM provides increased efficiency and quality to IP networks. It is a dynamic circuit-switched technology that provides transport between routers through channels and enables high-speed optical transport. With DTM, each channel has dedicated bandwidth and forms a dynamic route between the sender and receiver, passing through the routers along the path. Quality of service levels are established quickly and maintained from origination to delivery. Routers along a channel's path easily pass data from one link to the next because no address information must be checked. To optimize the use of bandwidth, transparent channels can be set up on demand in steps of 512Kb/s up to the full fiber capacity.

While transmission networks have been able to transport video for a number of years, the quality of the video has suffered because network resources were inefficiently used. The poor video quality has slowed the adoption of video-based products offered by network operators. And inefficient network resources did not give network operators the financial incentive to build video-rich networks. However, with advances in DTM, the industry is presented with a viable solution for efficiently delivering high-quality video while using maximum bandwidth.

DTM's ability to maximize bandwidth provides broadcasters with substantial efficiency gains. Because channel size matches the payload, channels can be built and added as needed. And because existing channels are not adversely affected by the addition of new channels, broadcasters are often able to add two HD channels to their existing infrastructures without adverse affects.

The equipment allows HD cameras to plug directly into a fiber-optic network, so content can be transported in an uncompressed format to product facilities or straight to the studio. Additionally, by transporting real-time video streams in the ITU-R BT 601 format, production units can be connected to video-production networks on-demand, providing an opportunity for new revenue-generating services. For example, TV producers could do real-time editing from different locations or set up film banks and sell raw film material to other TV producers.

Imagine a world where real-time HD video is available on demand from virtually any live sporting event to any home. It's a scenario that changes the role of every player along the media chain. This is the promise of the fiber already in the ground, when IP and advanced technologies like DTM come together. With DTM, the promised “it's like I was there” experience of HDTV is a current reality, deliverable to billions of viewers in the short-term here-and-now.

Mahmud Noormohamed is vice president of business development for Net Insight.