What's The Best Bandwidth Platform For Video Conferencing And Multi-Media Functions?
Designing the "perfect" delivery infrastucture for video conferencing and multi-media functions can seem complicated, labor intensive, time consuming, and costly. But what's most important is the evaluation process you would use....what you would likely choose (e.g. T1, DS3, OC3, Ethernet)...and why. Have a plan for that....and all will fall into place.
First, you'll need to ask and answer the following questions:
•What type of content do you want to use for your Video Conference system. Is it going to be just video of talking heads, or do you need other video and audio sources as well?
•How many participants would you have?
•Do you need to speak to multiple locations at one time?
•Do you want a dedicated room or a mobile solution?
•What is your existing IT/Network infrastructure?
•Do you have other Video Conference systems? If so what types?
These are the types of questions that must be asked when you're in the process of designing a Video Conference solution. The answersthen drive the selection of a Codec, it's hardware/software options, and which manufacture best fits the needs.
If the correct front end product is selected it makes it easier to integrate into a new or existing IT/Network infrastructure.
Rather than focus on the underlying physical topology used, concern should be toward finding a network provider that can deliver the features needed to support the application. It really makes little difference what the physical or link layer is. The provider will likely manage the CPE anyway and will give you an Ethernet handoff for your network. What you are concerned about is the network layer and more specifically QoS (Quality of Service).
To support streaming media such as voice and video, there must be a QoS mechanism at the network layer that will guarantee low latency and jitter. Converged network products offered by carriers today have this, usually using MPLS (Multi Protocol Label Switching) technology. This allows the customer to designate which packets get certain preferential treatment as they are sent along the network path.
As an example, consider a network that would deliver voice or video, VPN, and Internet access. Perhaps the application running over the VPN provided a core business function. In this case, data would be classified on the edge of the network to provide the greatest functionality. Voice and video packets would be categorized as latency and jitter sensitive, the VPN packets would be categorized as drop sensitive, and the Internet traffic as best-effort. These attributes will be honored within the carrier's network. The net effect of all this is that your VoIP phones and video will perform without noise or other problems and your Internet connection will just slow down a bit when you take the phone off the hook or put demand on the VPN.
Legacy networks have a hard time implementing QoS functionality due to the constraints of the older design and hardware (e.g. older broadband cable networks). Almost all networks built today can provide QoS but some designs work better than others. Always get references of other customers using a similar application to be sure you will be successful.
When projecting bandwidth requirements for the design, remember the human factor. If bandwidth is available, people will use it. The general trend I've seen is that usage doubles every year. Of course, adding additional networked applications can make it grow even faster.
In summary .... it all comes back to your requirements and setup backing up the deployment , such as :
•How many concurrent users
•Type of Service to be deployed
•Connectivity technology for each peer participating.
•Required bandwidth (with room for reasonable growth) to fill the gap.
If you are just looking at a couple of talking heads with little side data, you can get away with a few hundred Kbps per connection. However, if you are looking at medical consultation during a surgical procedure in hi-def, then 20-30 Mbps or more per connection may be required.
Next, what else is going on in your network? If this is a converged network (and that will ultimately be the way to go), then this video is competing with voice, other videos, data and who knows what else. What techniques are available to manage the data on the network? Even a gigabit Ethernet can get swamped if there are lots of HD video flows on the network.
Finally, what level of quality is being demanded by the users of the system? While you can do a "video" conference with 128kbps, the video is quite poor, and any packet loss or data errors at all cause serious problems. If you want telepresence (the feeling of being there) then you demand 1-way delays less than 250 MS, which limits compression (group of frames) options and increases bandwidth requirements.
The bottom line is that there will be a large tradeoff between capability, quality and bandwidth requirements. Once those tradeoffs have been decided, then you need to decide if your existing network infrastructure can support the trade-off decision or if you need to run a separate network to support the video. Considering that a separate network is very expensive, you then must decide what technology is required on your converged network to support this video application.
So, the question is not what is the best platform? Instead the question is what do you want to do? And the answer ..... is how to do what you want to do. Everything else -128K, 512K, 20 meg, SIP, H.263, H.264, JPEG-2000, MPEG-2, and on and on .... are all just tools that can be used to derive a solution. But they are not the solution, until you decide what you may need.
For FREE assistance determing what bandwidth solution best fits your video conferencing/multimedia needs and network architecture simply ask for help here....
To learn more about the collaborative communication solution I always recommend when asked click on this link.... Conferencing Services