One of the first concerns raised by IT departments when they hear about IP video surveillance is its impact on the network, the amount of bandwidth consumed by the different IP cameras. It is an even more important factor when monitoring remote sites; in those situations, the wide area network (WAN) is often critical for business operations, and they cannot afford network outages caused by the video management system "wasting" precious bandwidth.
In Security Center 5.0, we redefined how we manage video streams, optimizing bandwidth consumption for systems deployed over multiple sites.
Understanding the network's topology
When Security Center is installed over multiple sites or network subnet, it is smart enough to discover the different networks and build graphical representations . Of course, the system administrator can refine the auto-discovered configuration, but the fact that Security Center understands the network's topology is extremely powerful. The system will always find the most efficient path between the client and the camera.
Imagine a situation where a major incident happens in a remote facility.If four different persons at the headquarter need to look at the same few cameras in real time, the majority of VMS and DVRs on the market will stream each camera four times, from the facility all the way to the headquarter. That means four times the same information streamed over the WAN. In some cases, the WAN will not be able to support it.
Security Center is smart enough to understand the limitations of the WAN and only stream each camera once from the remote site to the headquarter, then forwarding the video to the four workstations with OR without leveraging multicast.
If an operator monitors a live camera and needs to go back a few seconds to verify something, why should the VMS waste bandwidth (and time) downloading the video footage from the recording engine? This information has already been transmitted once to the workstation.
In Security Center 5.0, everything the operator sees in real time is stored in a secure cache. This allows real instantaneous and bandwidth-free playback. It's faster and totally transparent for the end user. If the operator requests something that is not available in the cache, the system downloads it from the server transparently. This mechanism is simple, but can make a big difference when monitoring remote locations. This is also very popular in casino environments.
We applied the same logic for video playback: once the operator reviewed some video footage, there's no reason to download it again from the server. During an investigation, the user will often replay the same sequences; Security Center downloads the recorded video only once, and then uses the local disk to replay the video. This is very similar to how YouTube downloads the clip only once.***Of course, the cache is secured and cannot be replayed by another user. Each time the Security Desk is started or closed, the cache is cleared.
Dynamic stream selection
The majority of recent IP cameras can compress video simultaneously in more than one resolution. When an operator is looking at 16 cameras on a screen, or when some cameras are displayed in really small tiles, it's a waste of network bandwidth and CPU to display a 4CIF, or even a megapixel camera.
Today, Security Center supports up to 5 different streams, and the Security Desk can automatically switch from a low resolution to a megapixel stream based on the size of the video tile. This behavior saves a lot of bandwidth, but also reduces the load on the CPU, by using only the high resolution stream when needed.
These 4 examples are just a subset of the innovative ideas we introduced recently with Security Center 5.0. We have several other impressive features coming in the next release (5.1) of Security Center, which we will cover in the near future.