Video conferencing is a challenging application for business networks. Not only does it require assured bandwidth per call, but it needs very low packet loss ( less than .1% over 5 minutes) and tightly controlled jitter (less than 40 milliseconds). Meeting these requirements on a converged network requires that Quality of Service (QoS) be implemented end-to-end. To prevent the oversubscription of bandwidth, video conferencing also requires either call admission control (CAC) or carefully calculated bandwidth for every call.
Business video conference environments
Typical video conference deployments now entail a combination of fixed room-based systems, laptop or PC based desktop video clients with webcams, and, increasingly, tablet and smartphone-based video clients. With the explosive growth of tablets and smartphones over the last few years, the challenge of preparing an organization to support video conferencing has become significantly more problematic. Fixed-room systems are in known locations, and PC-based desktop video systems tend to be used from a few known or predictable locations, such as conference rooms, office cubicles and external venues. Tablet and smartphone video is likely to be used spontaneously in much more random locations, due to the mobility of such platforms.
Desktop video conference support starts with fixed video deployments
In order to support mobile and desktop video conferences in an enterprise, organizations must first handle fixed video infrastructure. In effect, IT should prepare for video deployments in phases:
- Phase 1: Get the fixed video infrastructure running smoothly since this should be the most reliable deployment.
- Phase 2: Expand the scope of your video deployment to include desktop video clients.
- Phase 3: Once the other two phases are rolled out smoothly, you can address fully mobile clients, such as tablets and smartphones.
Calculating desktop video conference bandwidth
To prepare your organization for desktop video, the administrator should create a spreadsheet to calculate bandwidth demand at each site, using these steps:
- List endpoint types and number at each expected location. The endpoint type refers to the make and model of the video codec, for example, Cisco EX90, Polycom HDX and so forth.
- Identify call rate, or call speed, for each endpoint type. The call rate determines the maximum resolution that can be achieved on a video call. The higher the call rate, the better the video resolution but the higher the network bandwidth load. Call rate, historically, represents the bandwidth that would be in use on an ISDN dial-up call. For IP networks, there is a 20% overhead, so a 512Kb ISDN call would actually consume 614Kb on an IP network. We will make use of this fact in the last step.
- Estimate concurrent usage. For example, you will need to know how many desktop video conference systems at a site will be in use on calls at a given time. If there are six systems and a maximum of three will be in use, the concurrent usage will be 50%.
- Calculate bandwidth needed at each site. To calculate the bandwidth required for desktop video, multiply the number of each endpoint type by their respective call rates to get the total call-rate-based bandwidth. Then sum the demand from all endpoint types at the site. This number is then multiplied by 20% (to account for the overhead in the call rate) to determine the network bandwidth demand at that site.
Bridges and systems that run multipoint calls are special cases. Bandwidth demand for those systems will come from external as well as internal endpoints. Be sure and include an estimate of the external video calls that will come in to the bridge, as this will consume network bandwidth for the site.
Calculating bandwidth for mobile video systems
More about desktop video conference deployments
Will employee desktop video use cause network problems?
Learn how desktop video conferencing affects networks.
Understand desktop video conferencing interoperability concerns.
With the fixed and desktop video model now in place, mobile video systems should be considered. Venues and Wi-Fi access points need to be identified, and an inventory of mobile devices created. This will have to be an estimate, because an accurate inventory may not be available in light of bring your own device and guest Wi-Fi access. Another estimate will be needed for the maximum utilization of each venue, based on reasonable capacity of the space, and the percentage of occupants that may be using video at peak times. Peak usage will impact site bandwidth demand to join video conferences with other sites, and will also impact Wi-Fi capacity utilization.
The bandwidth demand model, with mobile device demand now included, can be compared to available bandwidth on a site-by-site basis. If needed bandwidth cannot be added, call rates may be adjusted downward or concurrent calls can be regulated by call admission control or by scheduling.
Additionally, Wi-Fi infrastructure should be reviewed at each site, comparing the demand model estimates to available capacity. Older 802.11a/b/g deployments that focused on coverage now need to provide capacity where needed. Update technology where necessary to 802.11n or 802.11ac. QoS should be implemented, and possibly CAC.
Ultimately, the organization's Wi-Fi experience will be compared to that of the local Starbucks, which typically has plenty of bandwidth and few issues connecting. With proper planning and careful administration, this can be the experience your users also enjoy.
For more information, learn what challenges to prepare for during a desktop video conferencing deployment.