Network video can be used in an almost unlimited number of applications; however, most of its uses fall under security surveillance or remote monitoring of people, places, property and operations. The following are some typical application possibilities in key industry segments:
– Retail
– Transportation
– Education
– Industrial
– City Surveillance
– Government
– Healthcare
– Banking and finance
Yet the ability to monitor and record locations and operations does come at a price to the network’s bandwidth. Over the years, network capacities or bandwidths have continued to rise as more and more applications move onto networks for efficiency and control. However, video applications deployed by security solutions are bandwidth-hungry functions that often need to be addressed to ensure Local Area Networks (LANs) and Wide Area Networks (WANs) are not significantly affected..
H.264
Over the past several years, H.264 has rapidly become the standard video compression method of the network video market. The benefit of H.264 is that provides good video quality at substantially lower bandwidth than previous compression methods. It does this with better error robustness and is designed to give lower latency, as well as better image quality with slightly higher latency depending upon the H.264 profile supported. In fact, H.264 has unique features such as better prediction of I-, P- and B-frames. It also provides an in-loop deblocking filter, which reduces some of the artifacting seen when using MPEG-4 video compression. MJPEG compression will still produce a better overall image than H.264, but the bandwidth efficiency of H.264 makes it superior.
As network speeds have increased and H.264 has become more prevalent, the concerns of supporting network video on the LAN have virtually disappeared. One of the primary reasons for this change can be attributed to H.264.
LAN video bandwidth
Today, most existing LANs utilise fast Ethernet standards. This means they can transmit 100 megabits (Mb) of data per second. The trend for new or upgraded LANs has moved to support gigabit Ethernet standards. These robust networks now have the capability to provide one to potentially 100 gigabits (Gb) of data per second, a significant increase over past fast Ethernet networks.
Video bandwidth consumption is affected by a wide variety of factors. As an estimate, one network camera might consume between 100 Kbps and 3 Mbps on a gigabit Ethernet LAN. In this example, the network utilisation would range from 0.01% to 0.3%for a single camera. If you were to have 100 cameras on the network the range would now be one to 30%network utilisation. This example may not create a problem for a gigabit Ethernet network, but it may be an issue for a fast Ethernet network when supporting 100 cameras. In this case, 100 cameras at 3Mbps would bring a fast Ethernet network to its knees!
Managing bandwidth challenges
Even with gigabit LANs, end users can still face bandwidth challenges supporting cameras on their network. The biggest challenge today is managing all the variables that can affect the amount of bandwidth transmitted from each camera. Below are just a few of the many factors that can affect camera bandwidth consumption levels:
– Low lighting
– Interference patterns (moire pattern)
– Motion
– Reflections and shadows
– Multiple streams
– Scene activity
– Frame rate
– Resolution
– Compression (MJPEG, MPEG-4, H.264)
– Default settings (manufacturers default to highest performance settings)
The challenge is that many of these factors are variable and can occur at anytime, making them very difficult to control. In most cases, a single camera will not be an issue but as more cameras are added to the network, problems can begin to manifest. Many of these issues are typically unknown until cameras have been deployed in the field, which can become the root cause of many cameras unknowingly being mismanaged and left to wreak havoc on the network.
As the system expands, there is also a wide variety of technologies that allow for the management of the overall bandwidth consumption, including:
– Event-driven frame rate
– Virtual Local Area Networks (VLANs)
– Quality of service (QoS)
– Constant bitrate (CBR)
– Data Distribution (unicast, multicast)
As more cameras are added to the network, the available amount of bandwidth will obviously decrease. One of the challenges of streaming video is that bandwidth consumption is variable and unpredictable, unless the cameras on the network are using a constant bitrate. Network video streams depend upon best effort delivery, which means the data can be lost or corrupted in transit. This can be a huge issue especially when video is considered a mission critical application. It is vital to ensure there is adequate bandwidth available – if video is not received the first time it is lost forever – there are no second chances. If bandwidth resources are running low, here are some of the issues that may begin to develop:
– Dropped frames – making video to appear choppy
– Artifacts – pixel blocking at an increase rate
– Image freezing or lose of connection
– Latency
– Jitter
Another consideration is meeting the bandwidth demands of client workstations. Network cameras are not the only culprits creating an increase in bandwidth consumption on the network. As network cameras push video across the network, client workstations are pulling video across the network for live viewing and review. Client workstation can vary in the quantity of cameras displayed, making them a potential challenge to manage. It is recommended that an administrator set the views and privileges for the various users of client workstations when ever possible.
WAN video bandwidth
As companies expand across geographies, more and more businesses have remote branches that their networks must service, creating and increasing the need for WANs. WANs present a challenge when it comes to meeting the bandwidth requirements of network video. In LAN environments connection speeds of 1 Gbps and 10 Gbps have become common; in WAN environments bandwidth is commonly less than 10 Mbps. Typically most locations operate on a T1 (1.544 Mbps) or less. So as companies and their network video applications grow more sophisticated, the thought process of how to best utilise WAN bandwidth must also be addressed. As the chart below shows, there are a wide variety of WAN connection speeds available. However, the financial burden to implement or lease high speed WAN connections makes it very difficult to justify.
The trend seen within the network video market has been to store video locally on the LAN and only pull it across the WAN when necessary. This is a trend that will continue until new technological advances solve the WAN bandwidth challenge.
How do we solve the WAN bandwidth challenge?
It’s too early to know, but new compression technologies may be the next breakthrough technology addressing video on the WAN. Currently, ISO/IEC Moving Picture Experts Group (MPEG) and ITU-T Video Coding Experts Group (VCEG) are developing a new video compression standard that is considered to be the successor to H.264. This new standard is sometimes referred to a H.265 or MPEG-H but is called HEVC (High Efficiency Video Coding) in the standardisation project. H.265 builds on decades of research and the fact that processing power continues to increase in our network devices, and uses more processing power to achieve more effective compression. It is estimated there will be a reduction of 50% in bitrate requirements for a specified video quality and a three-fold increase in the coding complexity compared to H.264.
Conclusion
Understanding the impact of bandwidth from network video applications can be an issue due to all the factors that can contribute to increased bandwidth levels. As LAN speeds have evolved and new compression technologies have become available, the concerns of supporting network video in a LAN environment have virtually disappeared. The challenge today has moved more towards supporting network video on the WAN. In most cases this is not a critical issue, since video can be stored locally or passed through the WAN when needed at reduced frame rates or resolutions.
With the rapid evolution of technology in the video surveillance industry, it is only a matter of time before we see new technological advancements address bandwidth challenges in the WAN environment.
Matt Powers is technical director for security solutions at Anixter and has over 11 years or sales, management and technical experience in the physical security industry. He serves as lead for security-focused testing at the Anxiter Infrastructure Solutions Lab and is a member of the Technical Services Committee for ONVIF.
