PoE delivers critical infrastructure for high definition IP video
In security and surveillance applications, ensuring maximum availability of high definition IP video, as well as cost effective deployment and running costs, are critical factors.
From security systems for public buildings to motorway traffic monitoring and city surveillance systems, high definition IP video surveillance is a rapidly growing market, where the quality of the network determines the reliability and quality of video delivery. Network redundancy is therefore a key factor in selecting the most appropriate solution. Ensuring that high definition IP video of critical infrastructure is always available, while minimising deployment and running costs are equally important. Network redundancy is therefore a key factor in selecting the most appropriate solution. Ensuring that high definition IP video of critical infrastructure is always available, while minimising deployment and running costs are equally important.
In order to minimise installation costs, the ideal solution is a full Power over Ethernet (PoE) deployment at video camera locations, while connecting to a full gigabit multi-ring topology that is capable of handling network traffic. At the same time, in the event of an incident, rapid sub-second recovery of the network is required in minimizing disruption to operators. The solution should also alert all the required operators as to what has happened and where in the network the incident occurred, so that any remedial action can be taken.
Factors to Consider: PoE/PoE+
For network hardware such as Ethernet managed switches and end devices (cameras), it is important to consider the IEEE standards relating to PoE, as some hardware manufacturers provide non-standard versions. The original IEEE standard for PoE devices is designated as 802.3af, which dictates a PoE maximum of 15.4W per port. However, this power rating may not be sufficient for the latest high power cameras, particularly the units that offer PTZ (pan-tilt-zoom) functionality. Many IP cameras now have integral motors and drives or other features such as fans or heaters. Fans, for example, may be required to prevent the inside of the camera dome from becoming 'fogged' with condensation due to temperature fluctuations. Some switch manufacturers may only offer units that satisfy the normal PoE power standard (i.e. 15.4W), which may not be sufficient. The latest IEEE standard is designated as 802.3at high power (or PoE+) maximum 30W per port, enabling more powerful PTZ cameras to be deployed.
Even if an existing network already has a variety of non-PoE switches installed, the IEEE PoE+ standard specifies that plugging in a non-PoE unit to the network will not harm this device as power is not sent until the switch (Power Sourcing Equipment) and the end device (PD powered device) have confirmed via an 'automatic system check' that PoE is actually required by the device. Another key benefit of deploying PoE/PoE+ switches is that they can be installed by network engineers rather than qualified electricians, further reducing deployment costs. Other advantages are that PoE switches offer a variety of manageable features, including:
• Power device 'keep alive' check - a managed switch periodically communicates with end devices in order to check they are still operating correctly. If an end device does not respond, the switch waits, then cuts off the power and reboots the end device before flagging this action up to the operator, who may wish to investigate further. Automatically rebooting a camera can save considerable time and costs by not having to physically send an engineer to the camera location to unplug/plug in the device.
• Power scheduling - systems can be set up to schedule provision of power to end devices, which can be switched off at certain times of the day when they are not needed. For example, a security camera in an office car park may be switched off between the hours of 8am and 7pm.
• Power priority - if there is a power drop over the network or emergency back-up power is required, the system can be set up to provide power to only the most critical end devices in the network.
Demanding new IP video applications require seamless video throughput. The end customer considers any breaks in video feed as unacceptable. This includes not only live monitoring of video, but also recording of video for later review, as well as live analytics for monitoring doors, areas and any other location of interest to security operators. For these types of applications, service level agreements may exist, which will specify that the network operator must provide a minimum uptime of, say, 95% or higher. Ring redundancy is therefore a critical factor. Ring recovery times may need to be down to sub-seconds (e.g. <5ms) for a bespoke 'all-the-same-switch' vendor solution, or <50ms for an IEEE standard interoperable system.
Today, network topologies therefore require more complex offerings than standard single ring designs. The world is moving towards multiple, interconnected ring topologies. For example, the new ITU-T8032 ring topology standard is designed to operate without 'broadcast flooding' to build the ring topology and which also allows a more unified interoperable approach from different vendors. This standard is designed to eliminate all the network conversation between devices to keep the network traffic to an absolute minimum.
Author: Menglei Zhang, Korenix Technology Published: 13 Mar 2014