As AV over IP becomes more and more prevalent in our industry, the one constant roadblock to success is communication with Enterprise IT professionals.  It all stems from the fact that our answers don’t always mesh with their questions.  Some of it is the terminology.  Most of it, however, is understanding what they need to know and how to express it.

First, we need to define the type of deployment.  There are three main types.  CONTROL over IP, COLLABORATION over IP, and VIDEO over IP (with audio, of course.)  Let’s take a look at the points that you will need to clearly communicate to an IT professional when discussing each type of deployment.

CONTROL:  Control systems that connect to a device on the IP network only send and receive short bursts of data at a time.  Therefore, bandwidth is never an issue.  As an example, in a single room, turning a few AV devices on, changing inputs, and turning the system off will send and receive less than 10Kb per hour.  If you have ten rooms, that is well less than 100Kb of data per hour.  That is less than saving a Word document or PowerPoint presentation. So, you can see that networking control systems will not affect a networks’ bandwidth capabilities much at all.  However, you will need to obtain a static IP address for every device that will be controlled on the network.  In order to be able to locate that device, we must know it's address can never change regardless of changes in the network topology.  Additionally, in cases of deployments with large numbers of networked AV devices, it can be advantageous to segregate these similar devices on their own vLAN.  This makes it easier to manage the network and provides room for expansion without affecting the mapping of the greater Enterprise Network.  In other words, you have your own network segment, apart from the rest of the network, so you can add and subtract devices as the need arises.

COLLABORATION:  Collaboration devices typically allow screen-sharing over IP, usually WiFi.  There are several features in wireless collaboration devices that affect network performance.  The biggest issue is video media streaming.  This is where the user streams a video from his or her device to the main screen.  It could also be where a user displays his or her device to the main display or requests to have the main screen displayed on their device.  File sharing is also possible on many of these systems, as is video conferencing.  While collaboration systems vary, the worst-case bandwidth usage could be as high as 300Mbps in a multi-station collaborative environment.  However, for most rooms, 8Mbps per system is a typical bandwidth requirement.  This will account for peak usage in a streaming application.  In a high traffic network environment, it is helpful to have IT apply Quality of Service (QoS ) on each system’s IP path.  In many situations, it will be much easier to have the collaboration systems segregated on their own vLAN so that it can be firewalled from the rest of the network.  This allows guests to have access to the collaboration systems without having to have an enterprise network account. There are plenty of manufacturer whitepapers available that provide all the detailed network configuration documentation on their specific collaboration systems.  It is also very helpful to remind IT, managers, that for optimal performance WiFi Access Points should be provided for every 10-15 connected users.

VIDEO STREAMING:  Video streaming can consume considerable bandwidth as we discovered above.  This is the main point of discussion with IT managers.  As a benchmark, 6Mbps per video channel is a figure that we use to allocate resources, as it is on the high-end of the practical scale.  Video streaming can be done using Adobe Flash and HLS TCP/HTTP or Multicast or Unicast RTP/UDP protocols.  Flash is quickly becoming obsolete due to the inability of many mobile devices to decode Flash streams.  Unicast UDP is easy, but is not practical except for simple point-to-point streams, because full bandwidth is required for each connected endpoint.  That leaves us with HLS (HTTP Live Streaming) TCP or Multicast RTP.  The typical bandwidth requirements for streaming are:

500Kbps – lowest required speed for streaming

1.5Mbps – recommended speed for quality viewing

3Mbps – Standard Definition video

5-8Mbps – 720p and 1080p High Definition

30Mbps – 4K Ultra High Definition

Forward Error Correction (FEC) is a big deal, especially when streaming HD and UHD video.  HLS does not support FEC.  TCP can recover from packet loss, but the video is still interrupted briefly.  That leaves us with Multicast RTP.  RTP with FEC “looks ahead” and corrects any data losses to prevent video interruptions.  However, this error correction comes with a price.  There can be a delay of about half-a-second latency.

While you may get push-back from IT managers when recommending Multicast RTP streaming, the advantages in the high-quality video will far outweigh the inconvenience of configuring a network for Multicast.  Watch One-hour in-depth videos On Demand Now!