Bandwidth Throttling

VMS Bandwidth Saving Technologies

Demand for video and improved resolution continues to increase in organizations

Beyond the core functionalities of recording and displaying video, one of the most important capabilities of modern Video Management Software platforms is the ability to reduce bandwidth required for transmitting live video. Using both optimized compression and bandwidth saving VMS technologies, remote and centralized viewing is becoming an industry-wide reality.

Bandwidth Throttling

Although not a true bandwidth saving technology, bandwidth throttling is a means to control the amount of data transmitted.

The VMS platform administrator is able to configure a limit to the amount of data transmitted from the VMS recording server to clients, typically configured in kilobits per second. It does nothing to modify or reduce the size of the data being transmitted, but instead simply prevents any amount of data over the configured limit to be streamed.

The feature allows administrators to accurately forecast the amount of data that could be consumed at a given time. This is most relevant when users of the VMS system may call up video remotely. Remote viewing would cause the video to be streamed over an internet connection, which may be shared by other network applications. Limited bandwidth availability may trigger a need to apportion available bandwidth in order to guarantee throughput for other uses.

One problem with bandwidth throttling for this use case is any user of the VMS platform would be affected by the limit that was set, not just remote users. If video were displayed on a separate client computer on the same LAN where sufficient bandwidth is available, the performance of local client viewing would be limited by this feature. To avoid local client performance issues, the configured bandwidth limit may need to be set too high to effectively prevent the demand for remote video viewing from consuming too much bandwidth over the internet connection.

When the configured limit is reached by a client, often some number of live camera feeds will freeze. If a relevant security event is occurring, this effect can cause monitoring professionals to miss the action, preventing a real-time response.

In practice, the feature is of limited benefit and other commonly available network services, such as Quality of Service (QoS), can provide the same functionality.

Multi Streaming

Multi streaming is much more advanced than bandwidth throttling. Multi Streaming works by pulling multiple streams of video from an IP camera simultaneously, and delivering the most appropriate stream of video, based on client viewing requirements.

Most modern IP cameras support delivering multiple versions of the same video feed simultaneously. Using a simple example, one stream can be delivered to the VMS at full resolution, 1080P perhaps (1920x1080), and a second stream can be delivered at a lower resolution, say VGA (640x480). Typically, the highest resolution stream is recorded to the VMS, so video evidence is always available at full quality.

If the client workstation is displaying the video feed in a small area of the screen, it may be unnecessary to transmit the full resolution.

In a case where VGA resolution is a higher resolution than the target display area but also lower resolution than the full-size video stream, the VGA stream would be transmitted. Because the VGA stream is still higher resolution than the target display area, video will be displayed at full quality. No additional benefit would be gained from sending the 1080P stream. In this case, the benefit is the reduced bandwidth of transmitting the VGA stream as compared to 1080P.

  • Although a VMS platform with Multi Streaming capability is likely to be much more optimized than one without, there are some limitations to its application:For there to be any benefit, the system must use Multi Stream capable IP cameras. Most cameras support multiple, simultaneous streams of video. However, there are often performance limitations of the camera that may require sacrifices with resolution or frame rate. If pulling many streams simultaneously, each stream may need to be set at a lower frame rate than desired because of CPU limitations on the camera.
  • Although Multi Streaming saves bandwidth between the VMS server and the live video display client, it consumes more bandwidth between the camera and the VMS server. Each additional stream from the camera to the VMS takes bandwidth on top of what would have been consumed by pulling only the stream used for recording.
  • When using digital zoom on a camera, the VMS will stream the highest resolution video feed from the camera to provide a greater number of pixels and therefore more detail. When digital PTZ is in use there are no bandwidth saving benefits from Multi Streaming.
  • Multi Streaming does not send the exact resolution needed by the client. The limited granularity can add up to substantially more data transmitted than what is necessary which translates to more bandwidth consumption. Using the previous example, if a client is displaying 16 cameras simultaneously on a 1080P resolution screen, the client only needs a total resolution of about 2 megapixels, or about 0.125 megapixels per camera feed. If each feed is being transmitted as VGA resolution the VMS would be sending a total of about 5 megapixels of video which is vastly greater than the client’s display requirements.

Multicasting

Using multicasting, data can be sent to multiple recipients simultaneously while using a minimum amount of bandwidth. With a multicast capable system, the sender will transmit data to the network, and each recipient will ‘subscribe’ to the data stream. Instead of the sender transmitting a unique stream of data to each recipient, the network replicates the data for each recipient.

Multicast works best when there is a single source of data and many recipients. An example of where this model is best used would be transmission of a live video broadcast, like a sporting event. Because many recipients want the same data all at the same time, Multicast conserves a lot of bandwidth by sending only a single copy of the data to each network segment where there are multiple recipients.

In a video surveillance system, there are many sources of data (cameras) and usually very few recipients (users). Think of a system with 100 cameras and a single video monitoring room.

Unfortunately, multicast has minimal to no effect in situations where there are only a few recipients, different users want different video streams, or users want the video at different times. For these reasons, Multicast provides a benefit only in limited circumstances for video surveillance.

In addition to its limited application in a VMS environment, multicast carries with it some additional planning and deployment challenges:

  • Transmission of multicast streams requires specialized networking equipment. Basic layer 2 switches may not have IGMP support required for multicasting. Network equipment with this functionality typically costs more than comparable network equipment, adding to the cost of the deployment.
  • Cameras (usually) need to be multicast capable, which limits camera selection when deploying a multicast enabled video security system. Most architectures involve the camera sending the multicast stream as opposed to the camera streaming unicast to the VMS then the VMS sending the multicast stream.
  • If improperly configured, multicast data broadcasts to all devices on the network. A stream will be transmitted to each device on the network segment and the device must process the data in order to determine what to do with it.
  • The deployment of a multicast enabled video security system requires significantly more planning and setup time than a unicast system. The network equipment, cameras, and VMS must be configured properly for multicast to function. This increases planning and installation time, and deployment may require specialized skill sets from installers all adding to the system cost.

If planning for a situation where a major security event occurs and multitudes of users want to see video of the event live, the Multicast may provide a benefit, but those circumstances are quite limited.

Transcoding

Using transcoding, bandwidth consumption related to streaming live video can be reduced by a greater amount than other methods detailed herein. Transcoding involves modification of the compression format, resolution or other elements of the video stream received from the camera.

As an example, if a 1080P monitor is used to display 16 cameras, the display resolution allocated to each camera is 0.125 megapixels. Any higher amount of resolution streamed will not result in increased image quality on the display. If each camera is 1080P resolution, then without some type of resolution reduction technique such as multi streaming or transcoding, the VMS server would stream 32 megapixels of video to display on a screen with only 2 megapixels of display resolution.

With transcoding, the VMS client application communicates the size of the area allocated for display to the VMS server. The VMS server then scales the video stream to that target resolution, therefore only the resolution that will be used by the client is transmitted over the network.

The same process would take place on any digital video surveillance system at the client, but with transcoding, the process takes place at a different location - on the server prior to transmitting video over the network, which results in the bandwidth savings.

Another effect of transcoding is reduced client CPU consumption. Because the scaling takes place at the VMS server, the client computer is processing less video. Furthermore, if there are multiple client computers connected and viewing the same camera, there can be system-wide processing consumption benefits.

Some VMS transcoding implementations provide a Digital PTZ Cropping feature resulting in additional bandwidth savings when zooming digitally. Using multi streaming, the highest resolution video feed is transmitted when users zoom in, so additional resolution is available to avoid pixilation in the image. Of course, when digitally zooming, not all of the camera’s field of view is visible, so transmitting the unseen areas uses bandwidth unnecessarily.

Using a transcoding implementation with Digital PTZ cropping capability results in only the zoomed area being transmitted for live display. However, the full field-of-view is always recorded and available as evidence for review and export. When compared to other VMS bandwidth saving technologies, transcoding provides unique attributes.

  • Greater bandwidth savings as compared to multi streaming due to more granular resolution adjustments.
  • Bandwidth savings when digitally zooming.
  • No specialized network equipment or configuration required, reducing setup time and network equipment cost.
  • Although the VMS server CPU load may be higher, client performance is increased and client CPU consumption is decreased.

Bandwidth consumption can make a key difference in how the system is used. Centralized monitoring allows for the possibility of proactive response to a security event, instead of using video surveillance only for gathering evidence. VMS based bandwidth saving technologies can often have a larger impact on bandwidth consumption than any other technology component, making this a key criterion for technology selection.

This article originally appeared in the February 2017 issue of Security Today.

If you like what you see, get more delivered to your inbox weekly.
Click here to subscribe to our free premium content.

comments powered by Disqus
  • Environmental Protection
  • Occupational Health & Safety
  • Infrastructure Solutions Group
  • School Planning & Managmenet
  • College Planning & Management
  • Campus Security & Life Safety