September 2004

Driving Adoption of IP-Based Video Telephony (continued)

FIELD UPGRADEABILITY
The last thing a consumer wants is to buy is an IP-based video phone that becomes obsolete after only six months. Fully programmable devices eliminate this problem, however, because they can be easily upgraded. Through the phone’s IP connection, the devices can upgrade themselves automatically by tapping into a master Web site on a regular basis and downloading the appropriate software.

Field upgradeability is crucial because the video market is in a constant state of flux. Those devices that can stay up to date will have greater staying power in today’s dynamic market. Devices that can’t adapt, however, should never be purchased in the first place. Today, most consumers expect to be able to download software upgrades for free, and while developers dislike this trend, intense consumer resistance to cost associated with this scenario indicates that the trend is likely to continue.

PERFORMANCE HEADROOM
Implementing new standards, more often than not, requires that the IP device has the flexibility to support the new standards, as well as headroom to handle the increased performance requirement. The new H.264 standard, for example, processes 41 motion vectors per macroblock — a significant advancement in comparison to the one motion vector processed under H.263. It takes significantly more computational power to process these extra motion vectors. Implementing H.264 in a H.263-based device, therefore, requires extra headroom or some as-yet unused processing capability.

Balancing initial product cost while still leaving enough headroom on the device is a delicate problem and requires a thorough understanding of the target market’s price sensitivity. It is important to recognize, however, that there is room for growth beyond simply implementing new codecs. Customers should also consider this capability when evaluating which system to purchase.

A device can vary in its display, to which it attaches and the security features it relies on to send protected audio and video over the Internet, which requires additional resources for encryption and authentication. Headroom is essential for enabling this type of functionality, both today and in the future. An investment in a system with appropriate processing power may prevent the purchase of a completely new system in the future.

INTEROPERABILITY & COMPATIBILITY
Flexibility in design is important because no one knows what the market will bring tomorrow. As broadband continues to reach more homes and small businesses, the market will see a need for more IP-based video client end-equipment, such as videoconferencing terminals and set-top boxes. With a diversity of endpoints, however, there will arise a variety of different data streams, protocols and codecs that devices will need to support. Devices that don’t support the most common and latest standards will be incompatible.

The end products shouldn’t be limited in capabilities. Customers want to link the video phone with several other devices, all of which may be using different video and audio formats. An essential aspect of supporting such capabilities is that clients have to be independent. They can’t depend upon some other node in the network to bridge between protocols or codecs. Not only would this increase latency and reduce video quality, it would be expensive to create and deploy such a device. Instead, each client must be able to independently handle the different formats internally.

Another angle of compatibility to consider is that of software drivers. For example, a user may want to use Meeting Room Manager Lite from Microsoft. If the device requires a unique implementation of the drivers for Meeting Room Manager and Windows Media 9, then the video phone won’t be able to support them until software is written specifically for that device. Devices that are designed on standard platforms can be upgraded immediately using the standard drivers that are freely available.

THE RIGHT UNDERLYING TECHNOLOGY
Providing these key elements — ease of use, reliability, full software programmability, field upgradeability, performance headroom, and interoperability — while keeping costs down requires a balance of integrated technology. In general, the more components a system has, the more complex, costly, and space intrusive it will be, however, these aspects may also help improve the longevity of the device. System-on-chip digital media processors provide all of the processing performance a device needs, as well as offering a number of the requisite peripheral drivers. Also, additional headroom in a single device can be a less expensive alternative to implementing extra headroom in an optional, second processor.

From a design standpoint, modular architectures simplify system design and reduce overall cost. A single base design, for example, can support a family of pin-compatible digital media processors, offering a wide range of performance and available headroom. Different modules attached to a primary board can “customize” the board for different applications, such as whether the device has its own display or drives a TV monitor. If a product does not need a particular peripheral like a universal serial bus (USB) port or liquid crystal display (LCD) controller, that slot can be left empty on the board. If there is only a single base design, this architecture is lower in cost than several specific designs because of higher volumes.

No one knows how existing audio and video standards will change and evolve, and fixed-function systems have no simple way to accommodate these changes. In the future, it will be the products based on devices that are the most flexible and that can adapt to the rapidly changing landscape of the video market that will be the best long-term investments for consumers. Prior to investing in a system, customers should carefully consider each aspect of the products’ design lest they find their cutting-edge investment will not outlast the next codec.

Page 2 of 2 [Back to Page 1]

Pradeep Bardia is worldwide marketing manager, DSP Video Solutions and Xiangdong Fu is senior software applications engineer, DSP Video Solutions at Texas Instruments. For more information, please visit the company online at www.ti.com.

If you are interested in purchasing reprints of this article (in either print or HTML format), please visit Reprint Management Services online at www.reprintbuyer.com or contact a representative via e-mail at [email protected] or by phone at 800-290-5460.

[ Return To The September 2004 Table Of Contents ]