Using Fiber Optics for SurveillanceBy Ethan Ace, Published on Dec 07, 2011
In using fiber for surveillance applications, there are typically two scenarios that users must address: backbone between two switches and directly to the camera. In this update, we'll address the separate concerns of each, and provide our recommendations.
General Fiber Limitations
In selecting fiber, the fundamental decision that must be made is whether to use multi-mode or single-mode fiber. Multi-mode is more common in indoor applications, as it is less expensive, but has a maximum distance of about 500m for Gigabit Ethernet, or about 2km for 100Mbps speeds. Single-mode, on the other hand, is more expensive, but can transmit tens of kilometers, depending on the hardware used, making it the only choice in some applications.
In these pricing comparisons, it is not only the cost of cable that is higher, but cost of connectors and equipment, as well, which widens the price gap. For surveillance, single-mode is likely not used, unless connecting cameras in a municipal infrastructure beyond the distances multi-mode can handle.
When connecting distant cameras, fiber is often the only option. In these cases, the most common way to connect the camera is to use a media converter at each end. This presents two potential challenges:
- First, unlike UTP, which can supply power over Ethernet, using fiber cable requires power at the remote end. In many locations where cameras are installed, such as on light poles, power is already installed, though may need a transformer to provide correct voltage. As we discussed previously, Berk-Tek also provides the OneReach system, which is essentially a fiber cable with copper power conductors to provide PoE at the far end.
- Second, and of lesser importance, is what to do at the near end of the fiber run. Some IT-savvy users are averse to installing media converters because they cannot be managed from the same interface as their switches. In these cases, it may be preferable to use a fiber switch instead of a rack of media converters. Generallly speaking, however, this is a more expensive option.
Very few cameras, and none from major manufacturers, offer direct fiber Ethernet connectivity. In typical situations, if media converters are used for cameras, far end units are installed in an enclosure at the pole or other location, sometimes co-located in a power supply enclosure. The near end units are installed in a rack mount kit, which provides a cleaner installation and easier access to cabling. Rack mount chassis also normally share a power supply, reducing the number of outlets required at the rack.
There are a number of common providers of media converters: Transition Networks, GarrettCom, ComNet, and EtherWAN are all oft-used. Users should expect to pay $100-200 for a media converter, or more if an environmentally hardened unit is required.
The second application in which fiber is common is connecting two switches in distant locations. Under 100m/290', UTP cabling may be used for backbone between switches without issue. Beyond this, however, fiber is normally used. While media converters may be used, as discussed above, typically they are not, opting for SFP transceivers instead.
Most professional-grade switches today are equipped with SFP (small form-factor pluggable) ports, which accept a small fiber transceiver which takes the place of the media converter. Typically, SFP ports are simply configured as uplink ports (if the switch is managed), and fibers are connected between the two switches. This greatly simplifies installation, since no external hardware is required, and the ports may be configured from the same interface as all other switch ports.
Users should expect to pay $75-150 for a typical SFP module, normally less than a media converter. Modules made for single-mode fiber and specialized applications (such as those used for SONET rings or single fiber applications) will be priced slightly higher, but not used in most instances.