By Brett van den Bosch.
Source: Hi-Tech Security Solutions Magazine : http://www.securitysa.com/59228n
When it comes to setting up a network for a surveillance installation, many of the same considerations that pertain to the IT domain still apply. However, it is generally more important to ensure consistent surveillance network traffic than it is for most other applications, particularly when that network might be relied on in an emergency situation.
According to Eurobyte’s Roberto Vizcarra, there are many factors to consider, “But the most fundamental and often overlooked for me is the need to identify the throughput requirement for your surveillance network, as well as when you break out of your network and go through the public domain (i.e. for remote viewing and alarm notifications). What I mean by throughput is knowing how much data is going to be used by each camera point through the entire network. Not tracking this often results in dropped connections and lagging streams, which are very common problems in IP surveillance setups today.”
Vizcarra offers the following tips for setting up an effective IP surveillance network:
1. Surveillance is critical, so always keep the network (physically or virtually) separate from other business communications. Sharing should never be an option.
2. Ensure your NVR can support the amount of bandwidth usage for the cameras provided (especially if you need to include third-party cameras on the NVR).
3. Map out the traffic flow, i.e., calculate how much bandwidth is flowing where. This will enable you to determine the correct media – e.g. what types of switches, and whether to use network cable or fibre – as well as what kind of Internet upload and downloads speeds you require for remote viewing.
4. If a camera is projecting to multiple NVRs, such as with a failover server or even a viewing station, the bandwidth requirement multiplies by the amount of devices the camera is projecting to. So, for example, one camera projecting a main stream of 2 Mbps of data will have to create 4 Mbps if that single camera is being recorded by two NVRs. To avoid bottlenecks use the multi-casting function on your switch.
5. Most routers (for breaking out the network) have priority or QoS capabilities, which allocates a certain amount of bandwidth (upload or download) to the system. Again, map out how many remote devices and alarm streams will be used (sub-stream data) and allocate the required bandwidth. Many new complexes and even crêches today are offering all their tenants and parents mobile apps to access surveillance viewing without actually catering for the bandwidth requirements, resulting in poor quality streams. This can be avoided.
Wireless technologies are attractive, mainly due to their ease of installation and low cost, but Vizcarra advises against taking this easy way out. “I want to use this opportunity to encourage people to avoid it,” he says. “Wireless should be seen as a last resort and only for those really difficult places to reach. Good wireless solutions should be done within your own property; since urban areas today are saturated with wireless transmissions across all frequencies, resulting in poor and dropped connections. Avoiding public spaces is a must.”
If wireless communication is deployed, clear line of sight is essential, since wireless transmissions cannot transmit through trees or walls. “There cannot be anything between the antennas,” explains Vizcarra. “Bridging through multiple antennas results in drastically reduced throughput and should be avoided too. All antennas and receivers have specified angles, distance, throughput and frequencies, so make sure their requirements are well met.”
Due to their ubiquity, cellular networks are often seen as a good option in certain applications, but it is important to bear in mind their limitations. “Cellular networks are used for surveillance purposes, but generally not for recording because of the cost. Rather, they are typically used for live viewing from a central location or via a mobile device. We often use cellular technology to transmit a live feed in mobile surveillance systems used in vehicles/trucks or even body worn cameras. These live feeds enable us to know what is happening on the ground (especially in emergencies), but the actual recording is generally done on the device itself.”
For enterprise applications, to keep data costs down when transmitting via devices through a cellular network, Vizcarra advises users to ensure the control room viewing ¬station, i.e., the software where the video feed is being live streamed from, has a blank screen monitoring option so it only livestreams or plays back the video recordings on request.
In summary, Vizcarra believes the key factors for creating a reliable surveillance network are understanding where your vulnerabilities are, and knowing exactly what you want to achieve from having the system. After that, it is important to scrutinise what the system is capable of. “An example of this is knowing the distance over which each camera will be able to recognise a face – this is often unknown by the end-user until he has the system. You should also know how much recording time you need, as well as other aspects important to your specific use case.
“Lastly, it is important to understand that surveillance systems do not deter criminals from entering premises that have poor security protocols. Pay attention to how you manage visitor entry and exit points, and how security responds when a crime occurs is very important,” he concludes.