Wireless Video Surveillance Tutorial

While wireless can uniquely solve certain challenges, it is far riskier to deploy and use than wired networks. As such, it is critical to understand when to use wireless systems and the key risks in designing such systems. If you use wireless networks prudently for video surveillance systems, the financial benefits can be quite significant. However, miscalculation in choice and design can result in significant reliability and scalability problems.

As a general rule, you should avoid using wireless networks unless wired networks costs are significantly higher than a wireless system. This is because deploying and maintaining wireless networks is far more risky and expensive than it is for a wired network. Wireless systems face much more serious problems that wireline networks do such as constrained bandwidth, signal obstruction, higher maintenance cost and scalability restrictions.

Let's review these key elements:

- How much bandwidth is available?

- How far can away can the wireless cameras be?

- How many cameras can I deploy?

Bandwidth

Wireless networks have far lower bandwidth than wired networks. On a wired network, bandwidth available for video surveillance can be easily 70 Mb/s to 700 Mb/s. On a wireless network, your available bandwidth is often no more than 5 Mb/s to 25 Mb/s. It is a dramatic and often overlooked aspect of wireless video surveillance design.

Wireless video surveillance usually has significantly less bandwidth the wireless system states. This is because the way bandwidth is calculated in wireless systems is the opposite of the more traditional wired approach. With a wired network, if you say you have 100 Mb/s bandwidth, this means you have 100 Mb/s going up and another 100 Mb/s going down. In a wireless network, if you say you have 11 Mb/s bandwidth, that is the total for both upstream and downstream. Some wireless systems are fixed to allow half the bandwidth for upstream and half for downstream. This is a big problem for video surveillance because almost all the bandwidth used is in one direction (upstream). Make sure your wireless system lets the upstream take up the whole bandwidth if needed. This is common with wireless systems dedicated to video but none in common commercial gear.

Environmental conditions often reduce the bandwidth further. Wireless networks are much more prone to effects from the environment than wired networks. Wireless networks will only achieve their maximum if the strength of the signal (signal to noise) is sufficiently high. If there are partial obstructions or if the antenna shifts slightly, the bandwidth from wireless systems can drop further. In our previous example, the 11 Mb/s wireless system only offers 5.5 Mb/s for streaming video. However, common environmental conditions can drop the bandwidth to 2.75 Mb/s.

Distance of Cameras

It is quite hard to set up multi-mile wireless links to video surveillance cameras. A number of factors including obstructions, frequency limitations, power limitations, and installation precision drive this. Note: this tutorial assumes the use of unlicensed frequency, by far the most common choice for deploying wireless video systems. If you are using licensed frequency, where you can use much higher power and ensure no interference, these issues are not as significant. However, obtaining licenses are expensive and time consuming so most application use unlicensed spectrum. The rest of the discussion assumes unlicensed frequencies.

You are constrained in how powerful your signal can be, significantly reducing the distance that you can transmit. The government restricts the power of your signal so that you do not overpower the transmission other users. However, this means it is much harder to push through obstacles and go greater distances. It also means that other users of the same frequency can reduce the bandwidth or block your signal. This is a major factor in the emergence of the 4.9 Ghz range for use in video surveillance projects as that range is dedicated to public safety.

Obstacles are very seriously problems for wireless video surveillance systems. Most wireless video surveillance system use frequency ranges that are easily absorbed by buildings and trees (2.4 Ghz through 5.8 Ghz). Practically speaking, you may want to transmit to a building 100 meters away but if another building is in between, the signal will be absorbed and the link will not be possible. You can and should use mesh networks to accommodate this but you must factor in the impact on the cost of the overall network.

Installation precision is key but issues can go wrong that may increase long term maintenance. Because of power restrictions, wireless video systems commonly use high gain antennas that increase signal power by concentrating it into a narrower area. This can help greatly in going longer distances or overcoming obstacles, however, it means the antennas must line up very precisely. If they do not, the performance of the system will degrade significantly. Also, if during the life of the system, either antenna shifts, the performance of the system could degrade 'out of the blue.'

Number of Cameras

The number of cameras on a wireless system is severely constrained due to bandwidth limitations and constraints on how far cameras can be placed. For any given wireless connection, the maximum number of cameras that can be supported is generally between 5 and 15 with the cameras being less than a mile from the receiver. Even 'VCR' quality video using a good CODEC will take about 1 Mb/s. This is significant when your are dealing with wireless links that may only support 5 - 20 Mb/s. The total number of wireless cameras can be increased by using multiple wireless connections or by combining wireless and wired networks.

A prudent practice is to use both wireless and wired networks with the wireless portion minimized to only the specific scenarios where deploying a wired connection would be cost-prohibitive. A typical example is getting a network drop in a building (either off the internal LAN or from a telco) and deploying a wireless link from the building to camera locations close to that building on poles or fence lines.

In any of these approachs, CODEC choice and resolution selection are key factors in the number of cameras that can be supported. In a wired network where 70 - 700 Mb/s networks are common, not compressing video heavily can work. However, in a wireless network, with 5 Mb/s to 15 M/bs available total, a single MJPEG standard definition camera could consume all of the available bandwidth by itself. Similarly, given the bandwidth constrains, megapixel cameras are especially challenges. Even with various optimizations, megapixel cameras can consume far greater bandwidth than standard cameras (assuming you use the same frame rate).

Conclusion

Wireless networks can solve applications where wired networks are far too expensive. By relieving the need for expensive construction projects, video surveillance can be deployed in places where it would otherwise be cost unjustifiable. However, wireless networks offer far greater challenges and risks in design and maintenance. As such a clear understanding of these elements and when to prudently use wireless systems will contribute to success wireless video surveillance systems.

Footnote: Today, we added a category for wireless video surveillance: http://ipvideomarket.info/topics/Wireless. Use the RSS feed for it to track news and articles.