Video conferencing is typically not a protocol that is accelerated under the common understanding of acceleration. It is UDP based and as such does not have the same limitations as TCP that would demonstrate a benefit with using acceleration. Video traffic needs to be provided QoS and bandwidth guarantees as a first step. If you are trying to solve issues with ghosting or choppiness, then QoS is the solution.
There are other types of video acceleration mechanisms that people might refer to, but these are not accelerated by addressing the performance impact of high latency:
- One is caching video objects like Youtube or Vimeo video files.
- The other is being able to convert a unicast video stream to multicast. This allows one video stream to come in on an internet connection (i.e. a live stream of the President addressing the nation) and then split that out to several viewing stations. It would then be one stream to many so the internet link is not saturated.
Caching of video files, content, and dynamic URL's like YouTube is a huge advantage to reducing WAN traffic and improving user experience. This technology should be employed on almost every WAN link with today's increased usage of video content on the internet and in collaboration solutions.
However, without QoS, caching is only a point solution. In order to have a complete solution to deal with video streaming, the following should be used together:
- Visibility to understand the behavior and utilization of the network.