Many of today’s electronic security systems are designed to operate on a communications/computer network. SDM NetWorkings can help you understand both the intricacies of communication networks, as well as how security systems fit into the networked world. It features excerpts from the “Technician’s Guide to Networking for Security Systems,” a book written by SDM’s contributing technology editor, David Engebretson. (See end of article for ordering information.)

When connecting video transmission equipment to existing data networks, a detailed knowledge of bandwidth and how to control it will help the security installation company best work with IT personnel, who are justifiably concerned that the introduction of video signals will degrade their enterprise network performance.

Video streams require a fairly steady amount of bandwidth to transmit enough frames per second, at a high enough quality, to achieve acceptable viewing and recording of images. Generally, this constant video stream will be flowing on a 24/7 basis, as most video security systems are configured for constant monitoring and recording. So we can consider the video streams to constitute a “base line” of consumed bandwidth on the network, leaving the remainder for the fluctuating needs of enterprise communications.

Potential growth of the number of devices on a network – either computer users or security devices or both – behooves security dealers to configure their video systems to use a minimum amount of bandwidth.

Options for Bandwidth Control

The number of ways that an installing company can tailor bandwidth usage will be limited by the feature set included in the IP cameras or video servers selected for the particular job. Not all products will have the complete suite of options.

The three generally available selections for bandwidth limitation are frames per second, scaling, and compression percentage.

Frames per Second (fps)

A simple method for reducing network video bandwidth usage is to reduce the fps rate being transmitted. The great majority of current IP cameras and video servers provide a simple selection for the maximum fps that a particular device will transmit onto the network. Many security dealers have experienced good success using 15 fps, which generally provides enough motion for viewing and recording needs. Rates as low as 7 fps are often used when a sizeable number of network video images need to be recorded.

If real time video is a requirement for a particular job, 25 to 30 fps can be achieved by reducing the sampling rate (scaling) and increasing the compression percentage, while using a reasonable amount of bandwidth.

Consider the following simplified equation:

Video Bandwidth = (scaling x compression rate) x fps

Reducing any of the components will reduce the overall bandwidth needed.

In general terms, scaling is a programmable selection in an IP camera or video server that defines how many lines of pixels of each image will be compressed and transmitted over the network. While 640 x 480 provides the best image, and the largest file sizes for transport, a video server could be set for 320 x 240, for example. Choosing one of the lesser scaling settings greatly reduces file size. By reducing scaling settings, technicians also reduce overall compression/decompression time, which enables images to be displayed virtually instantaneously.

When viewed over the Internet, maximum frame rates will be dictated by the available upstream bandwidth of the client’s Internet connection. Adjusting the scaling downward will increase the number of frames that can pass through the connection to the remote viewing computer per second. Remember that upstream data is being sent and downstream data is being received.

Compression Percentages

Often an IP camera or video server will have an image quality selection, with descending choices such as superfine, fine, normal, and low. These options relate to the percentage of compression that is applied to the individual images, with the highest quality being the least compressed. Less compression equals better image quality and motion capture, but means larger file sizes, slower transmission speeds, and larger storage requirements.

Different manufacturers have developed their IP network cameras and video servers to work with specific compression methods. Typically a product will provide either JPEG or its motion adjunct MJPEG, or MPEG. While JPEG/MJPEG transmits whole and complete pictures, MPEG ships some complete and partial frames, which are referenced to the complete pictures.

In general, MPEG will provide smaller file sizes, while JPEG provides complete images that can be individually copied from the storage disk for transmission and viewing. It is important in the product selection process for the dealer to ascertain that the compression codec provided is compatible with the planned storage and viewing equipment and software.

The amount of bandwidth that a particular camera/video server needs to provide adequate quality video can vary based on network factors and the type of images being viewed. Motion in a camera’s image requires more bandwidth to transmit than a still image, as the images are more complex. Cameras viewing a busy scene, or a camera that is programmed for preset pan/tilt/zoom functions will require more bandwidth than a stationary camera viewing a relatively “quiet” scene.

When sharing network cabling and transmission equipment with the enterprise, the astute security installer will most likely want to reduce image quality and fps to leave as much room on the shared network as possible for enterprise communications.

Sidebar:

Table

Image Range of Frames per second

Resolution Image File Sizes at 128 kbps

160 x 120 32-48k 3-4

320 x 240 64-96k 1-2

640 x 480 320-480k 0

This table shows some examples of the potential number of frames that can pass through a typical DSL connection providing a 128k upstream bandwidth, at different scaling levels. Notice that as the scaling level increases, the number of potential frames per second decreases. It’s important to know that in the example, the 640 x 480 scaling level will transmit images, but they will be received at a rate of one frame every few seconds.