IP Video Installations: Solutions for Common Problems
September 1, 2006
As video surveillance systems are being increasingly connected to TCP/IP networks, issues arise that call for practical and innovative applications and deployments of devices. Culled from many training sessions led by the author, the following article explores some real-world issues that can be confronted by security integrators in the IP environment â€“ and how to solve them.
SWITCH PORT AVAILABILITY AND CONNECTIONSIf connecting IP cameras, encoders or decoders onto a clientâ€™s existing enterprise network, it is im-perative that the security installation company perform a careful survey of the telecommunication closets (also called Intermediate Cross Connects) where the IP-enabled security devices will be connected to existing network switches.
Letâ€™s look at an example where the security integrator has planned to install four IP video encoders in a specific telecommunications closet, but there are only two available RJ-45 ports on the switch â€“ all of the others are in use.
There are four possible solutions to this problem.
1. Buy a new switch. If rack space is available, a new switch could be connected above/below the current one, providing the needed Ethernet ports for connection of the video encoders. This answer may not be acceptable, as a new switch can cost thousands of dollars. However, if the IT department has been looking to upgrade its networkâ€™s switches, this might be an opportunity to obtain new equipment. This can make the addition of video security devices onto an enterprise network very attractive to IT management.
2. Use multi-input encoders. Some manufacturers provide encoders that can take four or eight camera inputs, and connect to the network via one Ethernet port. Although a multi-input encoder may be more expensive than the equivalent number of single-video input encoders, this may be a more cost-effective solution than obtaining a new switch.
3. Add an unmanaged switch. Unmanaged switches in four- and six-port configurations can be purchased for less than $25. This switch could be connected to one of the two existing Ethernet ports, and provide Ethernet connections for the four video encoders. This solution, although sound from a strictly functional perspective, will most likely not be approved by the clientâ€™s IT personnel. Enterprise-level networks usually operate using managed switches that allow IT to control and monitor a variety of network and user functions. An unmanaged switch doesnâ€™t provide the control capabilities desired or required by the typical IT department.
4. Relocate some encoders. As the encoders are going to be connected to analog CCTV cameras, itâ€™s possible that by re-surveying the job, the installation company may find another telecommunications closet that has available Ethernet ports that can be reached by extending the distance of coaxial cable used to connect to specific cameras.
As can be seen by these examples, one of the great advantages of IP networking is the flexibility of network connections. Once permission to connect the required number of video streams onto the enterprise network is obtained, it is imperative that security integrators carefully survey the job, looking for the specific Ethernet ports to connect the planned devices. Of equal importance are how the devices will be powered, and whether or not the security components will have backup power in the event of primary AC failure.
POE AND IP CAMERASPoE, or Power over Ethernet, is an IEEE standardized technology that provides 48 volts @ 350 milliamps DC current over the same two UTP pairs as are used to transmit and receive 10/100 Ethernet. In a typical scenario, a UPS is connected to a PoE-enabled switch, which provides short-term power to the connected devices in the event of power failure. PoE has become quite popular for powering VoIP telephone instruments.
There are several advantages when using PoE-enabled IP cameras on an installation. Separate wiring of the cameraâ€™s power is eliminated, along with trying to find unswitched AC outlets for plugging in transformers or power supplies. If the UPS connected to the PoE is of sufficient capability, the security camera can remain functional for some period of time after the start of a power failure.
Recognizing these advantages, a number of security camera manufacturers such as Axis, Bosch, and Sony are bringing PoE-enabled IP cameras to the marketplace, and this trend will continue.
Letâ€™s look at a typical scenario where a security integrator wants to install PoE-enabled IP cameras onto an existing enterprise network.
PoE can be supplied by either a PoE-enabled switch or a device called a â€œmid-span PoE hubâ€ that can be connected to an existing switch. While the security integrator might float the idea of replacing the existing switch with a new PoE-equipped network switch, this may not be attractive to the IT department. An alternative is for the security integrator to provide a separate mid-span PoE hub sized to provide powered Ethernet connections just for the number of cameras to be connected in that telecommunications closet. The security company can either provide a separate UPS to power the PoE cameras, or connect to an existing UPS with the permission of the IT department.
PoE is a powerful tool, reducing installation costs while increasing the viability of the security video system during power outages.
NETWORK VIDEO BANDWIDTH REDUCTION METHODS â€” "EDGE" RECORDINGWhile connecting video cameras onto enterprise networks provides a wealth of potential user advantages, the issue of bandwidth usage is not likely to go away any time soon. While IP-enabled access control panels, intercoms, and alarm transmitters only use network bandwidth for short time periods, video is often streamed 24/7 to monitors and network recording media, reducing the available bandwidth for other enterprise communications.
If video must be streamed across the network, security integrators can reduce video bandwidth by using MPEG-4 compression, setting frames-per-second rates to 15 or 7.5 fps, and using 2CIF instead of 4CIF resolution. If connected to a quality camera and lens combination, these settings often will provide high-quality video while dramatically reducing bandwidth needs.
Another method of bandwidth reduction is to only transmit video when motion is detected or some external alarm trigger is activated. This keeps the video stream off the network until there is some meaningful video to be viewed or recorded. Specific vendors provide this capability within their IP cameras, video encoders, and viewing/manipulation software programs.
A fresh concept to address the video network bandwidth issue is â€œedgeâ€ recording. Just as data files can be locally saved onto a network-connected PC or laptop without pushing the data through the network, video can be stored at the encoder or IP camera without going through the network. High-end video encoder vendors are providing USB ports on their devices, to which inexpensive USB hard drives can be connected and used for local video storage.
As most high-end encoders and cameras have multiple video stream capability, the various streams can be directed to different devices on the network at different times, or when alarms are activated. If the desire is to minimize video network traffic, the video can be stored at the edge, and only transported over the network when a user requests to review a specific cameraâ€™s recorded output. Based on programming and the edge deviceâ€™s capabilities, itâ€™s possible to transmit a cameraâ€™s stored compressed video data to a centralized network storage device at preset times, perhaps at 2:00 a.m. when the network is otherwise under-utilized. Once the video has been transmitted, the edge recorder may be programmed to overwrite its stored video.