14.3. Surveillance Systems and Videoconferencing
As you move to a converged network, you may want to factor video into your plans. Just as the converged network replaces the legacy voice network, it can also replace the legacy surveillance and videoconferencing networks.
14.3.1. Camera Surveillance
Most traditional surveillance networks consist of meters and meters of expensive, inflexible coaxial cable reaching from a low-resolution monochrome monitoring station out to a handful of analog surveillance cameras . The videocassettes used to record surveillance videos are bulky and prone to failure, and the quality of most surveillance video is poor.
With digital, IP-enabled surveillance cameras, a converged network, and good centralized surveillance software, you can eliminate all those shortcomings. Indeed, as you prepare your network for VoIP by implementing high-speed switching and quality of service, you're also preparing your network for video. Contrary to popular belief, it's surveillance, not videoconferencing, that ranks as the number one application for video in business.
Like VoIP endpoints, video surveillance endpoints (cameras) are fully self-contained, have a microprocessor, an Ethernet interface, and a variety of features. Some cameras have a web-based interface that gives you pan, tilt, and zoom control so you can point a stationary camera, zoom in and out, and so forth. More sophisticated cameras add infrared night vision, multicast streaming, and other cool features.
Video can eat up a lot of network bandwidth, 5 to 25 times that of a typical G.711 phone call, depending on the video codecs, resolution, and color depth employed. Some developers, like DIVR Systems (http://www.divrsystems.com) offer remote video surveillance solutions that ride on the same network as VoIP. In order to keep from having video swamp the network and break down phone call quality, your video surveillance apps need to play nice with QoS. This means making sure that surveillance traffic is treated with a lower priority than phone call traffic.
Video meetings are closer to telephony than video surveillance, but their requirements are a bit different. First off, while video surveillance needs reliable delivery to ensure that every frame of video is recorded, videoconferencing is more like a phone call; if a frame gets dropped here and there, no problem. So reliable packet delivery, a la TCP, isn't necessary. Also, the need for multicasting (that is, having many viewers watch a single video source) is more prominent in video conferencing than in surveillance. Finally, videoconference transmissions tend to be bursty and fairly shortunder a few hours usually. But surveillance video tends to be steady, and round the clock.
While not necessarily inclusive in VoIP, videoconferencing has implications for the VoIP network, just as surveillance video doesin QoS mainly . But, since videoconferencing is indeed a call-switching application, it is often implemented in, or with, phone systems. Some TSPs even offer videoconferencing service with special, camera-equipped screen phones. Packet8's VideoPhone plan is one such service. It allows Packet8 customers to place VoIP/video calls to other Packet8 customers.
Cisco's AVVID IP/VC solution allows video-equipped H.323 endpoints to participate in videoconferencing. SIPQuest Corp. offers a commercial SIP-based videoconferencing server that runs on Linux. Sony offers a SIP-based videoconferencing solution (Sony PCS) that works with any video-compatible SIP-switching server, such as NEC's Univerge SV7000. Technically, if SDP is used for call setup (as it is in just about every SIP implementation), adding videoconferencing to a call is merely as simple as supplying a SIP video phone on both ends.
As such, it's quite easy to create a videoconferencing solution with Asterisk. Start by enabling H.263 and H.261 video codecs in sip.conf . Video softphones you can try include LinPhone for Linux and FreeBSD, Windows (MSN) Messenger for Windows desktops, EyeBeam for Mac OS X and Windows, and iFon for PocketPC. SDP decides the parameters of the video channel during call setup.