We get questions daily from integrators and end customers asking about wireless IP video surveillance and I think some get frustrated when we answer, "well....it depends" and start asking 20 questions. Wireless is a fantastic technology and perfectly enables the use of IP video surveillance in those hard to reach areas of your campus but there are a lot of things to consider. So, in an attempt to reduce the number of times your hear that wonderful phrase here are some things to start thinking about...
- Forget the Hype: The first thing you MUST do is forget the datarate figures that manufacturers market on the side of the box. Manufacturers of 802.11a/b/g WiFi based hardware advertise datarates of 54Mbps and 108Mbps but your actual TCP throughput will normally be an aggregate of 15-25 Mbps usable (maybe less). However, some manufacturers will use proprietary hardware and software to increase the aggregate throughput to 30 - 35Mbps (point to multi-point) and up to 1Gbps (point to point). Look out for these if you need some extra juice.
- Line of Sight (LOS) it a Must: In ip video surveillance you need as much throughput as possible and in order to maximize throughput on your wireless deployment you will need LOS. Obstructions in your path mean major loss of system gain which normally translates to lower or NO throughput. Side Note: Many manufacturers will claim they have Non-LOS (NLOS) technology but what they really mean is that their products will work with mild Fresnel Zone infringment but not true NLOS. The only true NLOS wireless bridge is made by Motorola and is referred to as the PTP 300, PTP 500 and PTP 600. If you want more info on these let me know.
- Wireless is a shared medium: This means that all devices (wireless ip cameras, laptops, etc) connected to a single access point will share the limited throughput of that single device, making the access point the limiting factor in your design. A single standard wifi access point (discussed above) that has an average throughput of 20Mbps can only support 2 megapixel cameras streaming 10 fps of MJPEG and uses 8.58 Mbps per camera. If you do the math [20Mbps - (8.56 Mbps * 2) = 2.88 Mbps] the single AP only has 2.88Mbps of unused throughput left so it can only support 2 camera at a time.
- Reduce, Reduce, Reduce: The less throughput your cameras use the more cameras you can have on one AP at a time. To reduce throughput try using MPEG4 instead of MJPEG. Using the example outlined above and changing the codec to MPEG4 allows a standard AP to now support 6 or 7 cameras instead of just 2! Also think about archiving schedules. Stagger them so that all cameras are not archiving over the network at the same time. Sounds basic right? Well, you will be surprised.
- Use Sector Antennas: If you need more throughput than a single AP can give you use multiple access points with sector antennas to cover the same 360 degree area. If we use 3 APs coupled with 120 degree you can effectively triple your throughput!
There is more to consider when planning your wireless network like frequency reuse, RF interference, etc but this should hopefully get your started.
Thanks for reading!
- Ronen Isaac of Continental Computers & WLANmall.com
