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
Technology Business Applications
W O R K S H O P  
Videoconferencing: Making Your Bandwidth Count

  October 21, 2002
  By Mike DeMaria


>> continued from previous page

You Configure It Out
TOC Issue TOC
Printer Print full article
Printer Print this page
Printer Download as PDF
E-Mail E-Mail this URL
flameauthor Flame the author
 
  In this article
arrow
Introduction
arrow
You Configure It Out
arrow
I CU 2
arrow
Step By Step

The technical challenges of running a successful videoconference boil down to the setup, bandwidth, latency and appearance.

Setting up the videoconference is the most labor-intensive part. Most videoconferencing sessions take place over long distances, so stopping by a participant's office to set up his or her end isn't always possible. Therefore, the administrator will have to configure the setup remotely.

Every site needs access permission for an Internet videoconference. The trouble is, many firewalls are configured to block ports traditionally used for videoconferencing. You'll need to reconfigure the firewall to accept the videoconferencing session. Older firewalls may not recognize and support outgoing H.323 video traffic at all. And even if H.323 is supported, make sure your security administrator allows H.323 traffic to pass through the firewall.



NAPT (Network Address Port Translation), too, can cause problems with multiple videoconferencing users. If you have one videoconferencing user inside the NAPT, just forward all necessary ports to each of the users' internal IP addresses. It's a bit trickier with multiple users. Some NAPT boxes support H.323, but you can initiate calls only from inside the organization. You can't have someone behind the NAPT host a conference with people on the Internet.

For receiving calls to multiple or dynamically located users behind a NAPT, you'll need to set up a VPN tunnel that can handle NAPT traversal. The benefit here is the secured transmission of a VPN tunnel, because video over the Internet can be captured and replayed. If the videoconferencing software doesn't support encryption, you need to set up a VPN anyway. The downside is that this will add latency and overhead to the video stream.

Bandwidth is also a big issue in a videoconference. While a LAN usually provides more bandwidth than you need, that's not the case with the Internet. Sure, some users videoconference over a 33.6-Kbps dial-up link with AOL. But you only get about 0.5 frames per second that way. Reasonable conferencing can be done over a 128-Kbps ISDN line or a T1 link. The more bandwidth for a videoconference, the better.

Glossary
• DiffServ (Differentiated Services): A protocol for controlling traffic by classes so certain types of traffic get precedence.
• H.320: ITU standard for coding and establishing video connections. It also addresses point-to-point and multipoint sessions.
• H.323: ITU standard that extends H.320 to support packet-switched networks.


• MPLS: Multiprotocol Label Switching A QoS protocol that lets packets get forwarded at Layer 2 rather than Layer 3 between routers. It works with IP, ATM and frame relay networks.


• SIP: Session Initiation Protocol A protocol for creating, modifying and terminating multimedia and IP telephony sessions, defined in IETF RFC 3261.

Then there's latency. Before the videoconference, check out the packet round-trip time between conference sites during the time period the session is scheduled. Internet traffic varies throughout the day, so a video test at 2 a.m. will yield different results than one at 2 p.m. Does the latency remain relatively constant, or does it jump around? Large jumps and dropped packets cause poor video feeds and increased degradation. But the amount of latency isn't as big of a factor as is the consistency of the video-traffic flow. A connection with a small amount of jitter will handle streaming media better than one with a variety of latency times.

Solving latency problems can be difficult. Data sent over a LAN or consistent, stable WAN links offer much more control. You can use DiffServ (Differentiated Services), the emerging QoS standard, to prioritize video traffic on internal networks, assuming your network has enough bandwidth to support videoconferencing without sacrificing more mission-critical applications.

Over the Internet, you can apply QoS controls to the uplink point. But what happens to the data after that is anyone's guess. You can use traffic shapers such as Packeteer's PacketShaper or Allot Communications' NetEnforcer to dedicate bandwidth to the videoconference, or you can enable QoS tagging on the LAN side. These technologies also ensure other Internet protocols, such as HTTP and FTP, don't sap all bandwidth away from the videoconference.


start top  Introduction I CU 2 

Research and Reports

Hypervisor Derby
August 2011

Network Computing: August 2011

TechWeb Careers