IP Case in Point: Video Systems with Large Bandwidth Needs

How Choctaw Nation casino took a 300-camera IP video system onto their network


"Using a VCR for surveillance is like listening to music on an eight-track tape," said Breshears. "Using Digital Video Recorders (DVRs), which turn images into digital video, is like using a cassette tape while listening to music from a CD or mp3 player is more like IP surveillance - it is state of the art."

"Casino security requires advanced cameras that will capture the details of a card, a ticket, the number of chips thrown on a table or under one's hand," said Breshears of the system requirements for Choctaw's security team. "Network cameras produce the best image quality, resolution and clarity I have ever seen -- this is essential for our business."

Installations for the three casinos included all new digital systems in some areas while others used a hybrid system comprised of analog and digital cameras.

Getting the most out of an IP-based system

Casino environments are different from banks, retail stores and schools because they operate 24/7 and have light variations that challenge even the best security tools. There are low-lights, flashing lights and neon lights - all of which may distort the clarity and quality of the image.

The Choctaw Nation selected the AXIS 216FD camera, among others, for its advanced image processing that enables it to deliver crisp images in low lighting conditions. The camera has a built-in, 2-way audio capability including an audio detection alarm that allows for real-time communication with visitors or intruders. The installation also included other network fixed cameras and network dome cameras, also from within Axis' product suite.

Most of the cameras used in the casinos provided automatic iris control, allowing the iris aperture to change and maintain optimum light level to the image sensor. This feature is critical for capturing quality video images in the varying lighting conditions that casino and gaming environments often present.

Not only did the network cameras capture quality video, but they were able to do so without using excessive bandwidth thanks to adjustments in the configuration.

Large Bandwidth Installations

Network video products utilize bandwidth based on configuration. Transferring video over a network can, in some cases, overload the network causing problems with other mission critical applications. Since video surveillance was mission critical to casino operations, the design team needed to update and expand the network over which the surveillance video traveled. In this installation, the security team alleviated bandwidth concerns by purchasing and deploying new Cisco switches to support the expanded surveillance system along with a dedicated server.

New switches typically provide 1 GBit/s per port so that if the backbone is 10 GBit/s, close to 1000 cameras could run at highest resolution and highest bandwidth. So as long as the newer network equipment is installed, bandwidth is not really an issue.

Choctaw also adjusted camera configuration to reduce bandwidth use while maintaining resolution. For example, some cameras at 30 fps while others ran as low as 5 fps. Less frame rate equals less storage used on the servers, which can add up quickly if companies have multiple buildings to monitor as in the case of Choctaw Nation.

Regardless if a company has a dedicated network or not, there are steps an organization can take when designing an IP-based surveillance system to ensure that the network never becomes overloaded.

Bandwidth consumption can be reduced on each network camera by using different frame rates, video motion detection and MPEG4 compression. When evaluating resolution, for example, a high-resolution picture (4 CIF or Common Intermediate Format) contains four times as much data as a normal picture. A reduction in the frame rate by half, for instance 30 frames per second (fps) down to 15 fps, will reduce the amount of data transferred by half. Additionally, built-in intelligence functions such as video motion detection means a camera can be programmed to record and send images over the network if the video is worth recording. In essence, a deployment may have several cameras that are only recording video 10 percent of the time - again, saving bandwidth.