|
|
Video Compression is Integral Key for Triple Play

(Continued)
|
The video headend generally begins the process of transporting video to the home by acquiring content from satellite and terrestrial feeds. With digital content sources, operators have a choice to either perform a digital turnaround or re-encode the feed. Statistically multiplexed MPEG-2 Digital feeds are converted into constant bit rate feeds, and the transport streams from the satellite are streamed out to single program streams. This eliminates the need for multiplexing, a must in more traditional methods of broadcast video delivery such as satellite and cable. For analog terrestrial and satellite broadcast feeds the content is sent into the video encoder. All of this information is streamed out from the video headend using an IP, or in some cases, an ATM streaming device.
|
New Page 1
Sponsor

|
Building for Video-on-Demand
While high-speed Internet is certainly a vital part of the triple play network, this area ties in with on-demand services. These on-demand services can encompass a wide range of home networking services such as videoconferencing, multi-home gaming and peer-to-peer links for data transfer between homes. The most developed of these services is Video-on-Demand.
The crux decision of adding VOD to the video headend requires an understanding of the customer base’s peak usage. Considering the number of customers on the network as well as the current set-top box, the VOD system should be planned around peak usage. If it’s determined that the peak buy rate is 20% of all users (A typical peak might be people purchasing a movie at 8:00 on Friday night) the VOD server’s playout capacity should be established from those plans.
The next step is to determine where the VOD server will be placed within the network. If installed at the master headend, every VOD transaction will stream through the entire network. This requires a larger amount of backbone capacity to handle that peak transfer. Alternatively, the VOD server can be located closer to the network’s edge. While additional VOD servers may be needed to handle multiple transactions, backbone traffic is greatly reduced, saving on network capacity. Most VOD servers available today can be installed at either point in the headend, although some focus on mass storage and others focus on distributed architecture. Which version is more worthwhile depends on the size of the network.
Video compression
Video compression is perhaps the most integral key to a successful IP-based video network – particularly in a DSL infrastructure. MPEG-2 remains today’s standardized compression platform and typically encodes services anywhere between 2.5 and 3.5 Mbps per second. There are successful MPEG-2 based video over IP deployments throughout the world.
While MPEG-2 is viable for standard definition TV today, the efficiency of real-time encoding hardware vs the theoretical limits of the algorithm has reached a point of maturity where further improvements will be slight. The advent of advanced coding compression platforms like MPEG-4 Part 10 and Windows Media 9 will enable more simultaneous video streams per household, as the video can be encoded at lower bit-rates than MPEG-2 allows. With the lower bit-rates, the end user can receive three to four channels at a time with a longer loop length between the DSLAM and the home. This is particularly beneficial for rural operators who typically have some of the longest loop lengths. It will also be an important step in adding HDTV to the video portion of the triple play network. Initially though the headend and STBs supporting these advanced coding schemes will be more expensive than the MPEG-2 equivalents so the business case for deployment needs to be considered.
Approximately 10 to 12 Mbps plus overheads per channel are required to successfully offer HDTV over DSL. Since HDTV cannot be taken off-air as is possible with standard ATSC stations, as many as 15-17 Mbps may be needed unless technology is purchased to change the bit rate of those services. While VDSL and fiber-to-the-home can support those data rates today, it is not possible to support HD over MPEG-2 through a single line of ADSL, the more common DSL infrastructure without considerably shortening loop lengths. Although there is not enough data capacity in a single ADSL line without shortening loop lengths, there are migration strategies for providing HD over MPEG-2. “Line bonding” is a strategy that runs two ADSL lines into the home. The broadcast signal is multiplexed across the two wires to double the capacity. Therefore, a typical 10 Mbps ADSL line now offers 20 Mbps over two lines. Suddenly, there is enough capacity to support HD MPEG-2. The advent of ADSL2 goes one step further with the support of line bonding in its specifications.
<<
Previous page
Next
page >>
page
2 of 3
Send
us your response to
this article.
More
from the Blueprints: Telco Triple Play series:
|
|