Upcoming Events

Cloud Connect
Santa Clara
Feb 13-16, 2012

Cloud Connect brings together the entire cloud eco-system to better understand the transformation we're experiencing and promises to be the defining event of the cloud computing industry. Learn about the latest cloud technologies and platforms from thought leaders in Cloud Connect’s comprehensive conference.

Register Now!

More Events »

Subscribe to Newsletter

  • Keep up with all of the latest news and analysis on the fast-moving IT industry with Network Computing newsletters.
Sign Up

FEATURES Continued

Video Servers: Live From Your Network

by Todd Tannenbaum and Andy Covell with Dave Brown

To view theReport card.
The unicast approach can support VOD, while multicast technology can do only NVOD through repeated retransmission of multicast programs. Multicast transmission is more efficient, serving multiple users (often with better quality) from a single video stream. The unicast approach eats additional bandwidth for each additional user.

To ensure bandwidth for IP-based video transmission, you must test and monitor video traffic on your network and use that information to project potential peak demand--a difficult undertaking. Some client/server solutions help you manage video streaming on a packet switched network by means of a throttling mechanism that prevents video data from flooding the network and intruding on other network applications.

Though you can work to ensure quality video on your intranet, the larger Internet is beyond your control and delivery to Internet clients is risky. Even with a very fast network connection, remote users frequen tly will experience dropped audio and freeze-frame video when the network gets overloaded (for more on this see "Streaming Audio and Video on the Internet," August 15, page 110).

As digital video creates demand for greater bandwidth, most network managers eventually will opt for ATM as a WAN solution. But on the campus network, other backbone technologies, such as Fast Ethernet and FDDI (and eventually Gigabit Ethernet), will suffice, and many will opt for the relatively low-cost and evolutionary approach of these fast frame-switching technologies in the short run. The same is true on the LAN. Switched Ethernet and Fast Ethernet will dominate in the near term. Keep in mind, however, that when it comes to video, a 10-Mbps pipe may not suffice in the long run. For instance, MPEG-2 is designed to go well beyond 10 Mbps; you won't get high-quality (HDTV, for instance) MPEG-2 encoded video over a 10-Mbps private Ethernet.

On the home front, ISDN is just making inroads as a replacement for plain old telephone service (POTS) and modem connections. It runs at 128 Kbps, with both of its "B" channels bonded--just enough for decent low-bit-rate video using H.261 compression. But ISDN's days may be numbered. The cable modem is set to provide a shared, frame-switched connection that will run TCP/IP at much higher data rates than are possible over ISDN. It looks like the telephone companies will work to deploy High-bit-rate Digital Subscriber Line (HDSL) and Asymmetrical DSL (ADSL) a soon as possible--which will offer a 10BASE-T interface in the home--to be followed by Very-high-bit-rate DSL (VDSL), with its higher bandwidth ATM connection. These higher-bandwidth connections, too, will target Internet access in the initial stages, further accelerating evolution of Internet video.

It may be some time before bandwidth to the home is in the multi-Mbps range, which will be necessary if high-quality networked video to the home PC is to become a reality. Likewise, video on the enterprise network will take time to gain momen tum. But the tools are available to begin exploring this new frontier of technology, and your efforts now will be rewarded when video on the net eventually breaks new ground.

How to Enjoy Intranet Video On a Shoestring Budget

For many emerging network video applications, video quality is not paramount, but cost considerations are. Video for corporate communications, training, surveillance or manufacturing-process monitoring, for instance, may not require 30-fps, full-screen video, while implementation must fit within a very tight budget. Sound like what you're looking for? Then one of the low-bit-rate intranet video products reviewed here--Enhanced CU-SeeMe from White Pine Software, VDOLive 2.0 from VDOnet Corp. and IP/TV from Precept Software--may be a perfect fit.

We were especially interested in these products' suitability as a one-to-many video broadcast solution. Though the three employ different approaches to network video delivery (see "Low-End Video on the Internet," page 65), they share some important characteristics--all use TCP/IP, can run entirely on Intel Corp. microprocessor-based hardware, are inexpensive to implement and are limited in their ability to scale up in terms of number and quality of video streams (unlike the high-end products we tested).

We cranked a lot of video through these systems and played with the various capture, compression and administrative settings to get a handle on their strengths and weaknesses. We also created two video clips for comparison: one a low-action, talking-head clip created using a consumer-grade 8-mm camcorder, the other a VHS copy of a high-action sequence (Syracuse University's first touchdown pass in the 1988 Sugar Bowl pitting Syracuse's Orangemen against the Tigers of Auburn University). We encoded and captured these, then streamed video from the resulting files. We compressed and transmitted these clips on the fly--a technique referred to as live encoding.



Big, Bad E-mail

Return To The Table Of Contents


Updated October 8, 1996


Research and Reports

Hypervisor Derby
August 2011

Network Computing: August 2011

TechWeb Careers