Designing your networks for IP video

So you've made the decision to go with an IP Video Management system. How do you design your network to handle all of the video from the IP cameras? Here's what to think about:

A. Bandwidth available: If you're installing the new camera system on an existing network (not recommended), how much overhead do you have to devote to the new system?
Some common guidelines include:
1. A maximum of 5 Mbps on each network port
2. Not more than 100 Mbps per uplink
3. Not more than 500 Mbps per NVR server port
Remember that to transmit IP video images that are measured in bytes per second across networks that are measured in bits per second, you have to multiply by 8.

B. Power and signal feeds: CAT 5e and CAT6 cabling, already installed in the building, has distinct advantages. Connecting a Power over Ethernet (POE) switch port to an IP camera allows both the signal and the power to be transmitted across the cable. This can result in saving thousands of installation dollars since the additional power cable does not have to be run. Next, determine what the powered device will require from the switch. This is referred to as a POE class, and there are four classes, zero through three, each requiring a different amount of power draw in watts. The design engineer can also figure in uninterrupted power supplies (UPS) back in the data room. When the power is lost, the camera is still recording and video is still available from PCs with UPS backup. It is now easy to figure the power budget by using the following formula:
Number of cameras x power class (in watts) = power budget

C. Traffic Segmentation: There are two different ways to divide or segment traffic on an Ethernet network. One is to physically separate the IP camera network from the company’s production network by running separate network cables and installing new switches. This is by far the most expensive way to segment traffic. The other way is to program Virtual Local Area Networks (VLANs). This method utilizes the existing switch and separates specific ports into their own smaller networks.

CCD vs CMOS

The image sensor is the camera component which captures light and begins the process of turning it into a digital image. There are two types of security camera sensors: CCD (charge coupled device) and CMOS (complementary metal oxide semiconductor). Which of the two you choose depends on your application.

With a CCD sensor, every individual pixel's charge is transferred through an output node, which is converted into an electrical signal. The signal is then buffered and sent as an analog signal. Because the pixels are devoted to light capture, the image quality is usually pretty high.

With a CMOS sensor, every individual pixel performs its own charge-to-voltage conversion, and the sensor performs amplification and noise-correction. The sensor also includes digitization circuits, which allow the component to output information in a digital format. Because of the complexity of this design, the area devoted to light capture is reduced. And because each pixel must perform its own conversion, uniformity (image quality) is lower.

Both technologies have their place in the marketplace. If your goal is to have good low-light-level surveillance, then a CCD camera is the best. If there is little to no light, CCD sensors used with IR illuminators produce an excellent picture. If indoor lighting is adequate, and megapixel quality is your goal, then CMOS technology may be your best bet.

Make sure you test your camera with the lighting that will actually be present in the field of view. Do not rely on manufacturer’s data sheets for the minimum useable picture. This figure is often subjective, and your customer may not find the results acceptable. CMOS cameras are more prevalent in the IP network world. These cameras are typically equipped with built-in Web Servers, which enable remote viewing of video from other locations, by using IP network protocol.

In conclusion, CCD sensors are the best for outdoor applications, low light scenarios or when a higher quality image is required. CMOS sensor cameras should mostly be considered for indoor applications, though with adequate lighting conditions, they may also work well outdoors.

School Visitor Management Software

Your secretaries might be getting tired of issuing paper badges to visitors and signing them in with pen-and-paper logs. Do they want an easier way to track who is inside the school and make sure everyone is evacuated during an emergency? Are you sure you aren't giving access to suspect individuals?

There's a system to address this. It's called Secure Visitor Management Software (SVMS). SAGE isn't an approved vendor, but the technology is definitely interesting enough to highlight here.

It's very simple to operate. A visitor comes into the lobby and has his or her driver’s license scanned. The SVMS converts the information on the driver’s license to a defined format and prints out a temporary badge allowing access to certain areas of the building, using your existing access control software. The software can even check a national database to ensure visitors aren't on a convicted offender list. (If they are, it sends an email and alerts administrators that he or she is attempting to gain access to the school.) It can even track ongoing custody battles between parents.

Read more about SVMS

Designing Access Control

In a traditional wired access control deployment, a door controller is connected back to either a (non-IP) network controller or a building controller. The door controller typically handles two card readers, so it can regulate one or two doors, depending on whether one door is using a card reader for both entrance and exit. It would typically control three devices: a door contact (which tells the system if the door is opened or closed), the door lock and a card reader. Above the door frame would be a powered junction box, which powers each of the devices.

Security systems are now being designed across networks, and are using TCP/IP protocol to communicate to remote devices. Access control is no exception.

In IP-based systems, there is a direct network connection to either the door controller or the card reader, with the network connection providing low-voltage power, typically called Power Over Ethernet (POE). POE powers the door contact, the lock, the card reader and the request to exit (REX) device.

One of the most critical issues in designing access control or any network infrastructure is ensuring the right power source equipment is selected for the job. Many POE network switches do not have full POE capability to every port. When specifying a certain POE switch, make sure it provides the maximum POE wattage of out every port. The current standard is 15.4 watts; however, a new standard of POE plus is being developed, which will provide a maximum wattage of 30 watts.

One excellent example of an IP-based access control system is MAXxess’ netEDGE door controller. The NetEDGE is a high-performance, single-door controller that features a Linux operating system. The Linux OS is embedded in the netEDGE and greatly enhances the reliability and capability of the security management system. In addition, it provides several features to improve performance in any size application.

Because the netEDGE utilizes POE, separate power supplies and multi-door controllers are no longer needed. When utilizing POE for the controllers, readers, lock and REX power, a reduction in installation costs of at least 25% can be realized.
For more information on MAXxess’ netEDGE products, go to http://www.maxxess-systems.com/ or email me at ddamron@sagetechs.com.