How to ensure high definition IP video of critical infrastructure is always available

Paul Boughton

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, says Santos Muro.

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.

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 order to minimise 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

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 normal PoE maximum 15.4Watts 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. 15.4Watts), which may not be sufficient. The latest IEEE standard is designated as 802.3at High power (or PoE+) maximum 30Watts per port. PoE+ therefore allows 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 these can be installed by network engineers rather than qualified electricians, which reduces 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 OK. If it doesn’t 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 – the system 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 on only 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.

Ring redundancy

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 per cent or higher. Ring redundancy is therefore a critical factor. Ring recovery times may need to be down to sub-seconds (eg, <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.

Hardware specifications

In security and surveillance applications, typically at least one managed switch in a ring is located in the main communications room, with the majority of switches installed at the point of the camera, for example, pole mounted at a main gate in a small junction box at the side of a railway line or motorway. Here, the main factor to consider is the temperature rating of the switch. Most switches operate between 0°C and +40°C. However, over the past few years, temperatures across Europe and the UK have been pushing towards new records, with summer temperatures in the high 30s and winter temperatures as low as -20°C, particularly in the more rural locations. In these environments, ruggedised switches will need to be deployed.

Designed from the component up, these switches typically offer operating temperatures of -40°C to 75°C, although this will vary from one switch manufacturer to another. These switches have typically undergone rigorous specialist testing to ensure that they can operate reliably at these extreme temperatures.

When selecting a suitable switch, vibration is often overlooked. This is particularly important if the switches are installed close to a motorway or railway line. The vibration from HGVs or from trains travelling past 24/7 is often sufficient to break any standard device within a short period of time. For very harsh environments such as waste recycling plants, where highly corrosive gases occur, or for very high humidity conditions, optional special coatings can be applied to switch components to prevent parts actively corroding. As well as switches, any Small Form Pluggables (SFP’s) used in the application (e.g. optical transceivers, fibre optic modules) also need to be industrial-grade rather than office-grade devices, otherwise they are likely to fail.


All managed switches typically have a full software suite that allows for complete configuration into any network topology. As with all software, patches and bug fixes are required from time to time. These are standard across all reputable switch manufacturer ranges and are typically provided free-of-charge to customers. Allowing software updates also offers the opportunity of providing the customer with new additional features to an existing hardware platform as these are developed over time.

Santos Muro is business development manager at Korenix UK, Redditch, Worcs, UK.