Maximising Bandwidth During COVID Contention
During the COVID-19 crisis with more people working from home there’s been a major upswing in network contention as providers battle to meet demands with underfunded infrastructure.
The challenge for security integrators and end users managing enterprise electronic security solutions incorporating access control, alarms, automation and CCTV is trying to balance performance and bitrate in real time in wildly variable network conditions.
Storing footage in local servers or NAS locations, storing at the edge using in-camera micro-SD cards and the implementation of recording on motion can all assist in serious security applications. It’s also important for end users to be across performance in real time, if CCTV is mission critical.
Things are a little different with alarms and access control, thanks to storage of events at the door controller, or the local controller, as well as signally of alarms and other system events to a remote monitoring station.
Depending on network topology and the severity of issues faced, it may be necessary to swing to a distributed cloud model, or to seek out alternative comms paths, including dedicated networks. Another option is switching to a distributed management and recording model until contention eases.
CCTV consultant Scott Myles of ICS Group says that with the pressures and time restraints placed on commissioning technicians, camera settings are often set to a global average, meaning bandwidth is in the hands of the Gods and the manufacturers, with bandwidth fluctuation dependant on camera view and the activity within the scene.
“Ordinarily, this may not be an issue, as with most modern networks having gigabyte throughputs and server hardware capacity meeting the desired outcomes, means that unless you’re streaming to the cloud or remote monitoring, a global one-size-fits-all setting may not be an issue,” Myles explains. “However, with older networks and for clients who want more bang for their buck, drilling down into the settings of each camera can have a profound effect on the overall performance of the solution.
“From my experience tweaks to camera settings have solved many performance and capacity issues, such as allowing additional cameras to be added to a solution, or delivery of longer retention periods. I sometimes think of the accumulation effect of bandwidth is much like my credit card, a little bit here and a little bit there, and in the end it all adds up.”
There are specific setting adjustments that can help, according to Myles.
“While each camera is different, they all have similarities such as resolution and compression, with many now having areas of interest settings that allow the installer to ignore sections of a camera view such as trees blowing in the wind that can affect bandwidth,” he says. “As these are brand specific and in most cases useful, there are some general settings that can be applied to achieve the desired outcome whilst reducing the load on the network.
“Firstly, I like to take the time to analyse the camera purpose in context to the position and what is trying to be achieved. Is the camera purpose for identification, or situational awareness? Has the field of view been appropriately set for the task?
“Getting the field-of-view setting right for the purpose of the camera is often overlooked. I have seen this many times where very little thought is given to FOV, where camera angles are set too wide for a given scene and objects such as TV monitors and unrelated traffic is captured that causes constant movement within the scene that unnecessarily increases bandwidth.
“In addition, I have often seen 4K cameras in close proximity of a single door with the primary purpose of identification, and although in this instance identification was clearly achieved, due to the cameras positioning and field-of-view, the resolution could have been dial back to say 1080p and still produced the desired effect while reducing bandwidth.”
Myles says the other setting that is commonly used to reduce bandwidth demand is compression.
“Although I’m not a big fan of overly applying compression to video, as the higher the setting the lower the quality, the use of an appropriate amount of compression and constant bit rates will certainly have an impact on bandwidth,” Myles explains. “As a general rule, for situational awareness cameras where identification is not the primary purpose you can get away with applying a bit more compression than you would with an ID camera. However for situational cameras I usually revert to area of interest functions and apply this to areas of the scene that are not critical, such as sky, trees and walls, etc, instead of overly compressing the video.
“In the end its all about balance, and if you have time I recommend experimenting with the settings to get the best results, If you can, remember to check again at night and different times of the day like sunrise and sunset. While all these settings are important, there is no denying the benefits of understanding the intended purpose of each camera, and setting up the FOV for optimum coverage and, if possible, removing any non-critical elements in the scene that will have an impact on bandwidth in the first place.”
According to Mark Shannon of BGW Technologies, there are a number of ways security integrators can tweak systems to ensure they make the most of bandwidth.
“Video consumes the most network bandwidth but there are very smart ways to minimise its impact, particularly when a lot of video conferencing is occurring during this restricted period for CONVD-19,” Shannon explains. “Simple methods like dropping the frame rates down, increasing compression and using more motion detection recording on cameras that are not in critical areas or cameras deemed to not to be critical can be implemented.
“Also, there are technologies cameras employ that may not have been activated. Features such as Panasonic’s clever smart-coding, Pelco’s smart-compression, or just changing the codec from H.264 to the more efficient H.265 can significantly reduce bandwidth.”
Brett Hansen, Milestone says the utilisation of many camera brands’ encoded data technologies allows security integrators to compress the data and stream it more efficiently.
“This solution is further enriched by Milestone XProtect, offering a spectrum of control options for frame rate, video quality, and recording metrics,” he explains.
“The recording of every pixel at full speed may not be necessary until there is some footage the operator wants to see. This is accomplished by recording at a much lower frame rate until an event of interest occurs, at which time XProtect is programmed to increase the frame rate at which the event is recorded, then throttled back down automatically for ongoing monitoring and viewing.
According to Genetec’s Lee Shelford, the efficiency of remotely managed CCTV solutions comes down to effective setup and configuration.
“If we know the bandwidth limits, we can program this into the system and add user priority, so the system will never saturate the link as it will send a message to the next stream requestor once it’s close to the bandwidth limit,” Shelford explains.
“For instance, the system will start dropping video performance delivered to lower priority users to make room on the network for priority users. However, even when juggling bandwidth restrictions, having communications and video encrypted is a must.
“Imagine having sensitive video pop up online after a man-in-the-middle intercepted the remote stream because it was not encrypted – the reputational damage alone for the organisation would be long lasting.”
Allen Hepburn of Bosch agrees there are multiple ways integrators can tweak systems to perform adequately in times of high network contention.
“Where supported, integrators should use H.265 compression and view/playback systems in sub stream,” Hepburn says. “For instance, Bosch’s H.265 compression allows you to save up to 80 per cent on bandwidth. Resolution of the sub-streams can also be lowered and analytic alarms should be set to a schedule to avoid unnecessary traffic.
Meanwhile, Tony Luce of Network Optix says functionality like adaptive scaling is the answer.
“Fighting with ISPs is kinda like throwing a rock at a bear,” Luce explains. “It’s not going to do much except make the bear angry. With Nx Witness VMS this problem shouldn’t exist – everything is automated via our adaptive scaling technology.
“In limited network environments Nx Witness desktop, mobile, and web interfaces automatically adjust the streams being served to keep connections stable and usable, regardless of where a user is from. If a system is connected via Nx Cloud and a direct NAT connection isn’t possible because of routing issues Nx Witness will automatically re-route streams to be proxied via our Cloud and a CDN that covers the globe.”
Andrew Cho of EOS Australia explains the way to manage bandwidth challenges is to be found in management systems.
“In our case, Digifort has been strong in bandwidth management for a very long time,” he explains. “It has been improved constantly due to increased traffic generated by higher-resolution cameras. Most importantly, this tweaking is inherent within Digifort and it’s very easy to setup.
“And when using a web client, integrators should select a solution developed with HTML5 H.264 instead of MPEG, which is heavy on bandwidth, as well as being high latency.”