While most users in an enterprise network want mobile devices connected over Wi-Fi or cellular wireless networks, a large number of users still have cable connections. Generally, these users require high bandwidth and processing speeds and/or secure connections.
Traditional structured cabling would have them connected to an office outlet with a patchcord and the outlet connected to a switch in a telecom room over category-rated, unshielded twisted-pair (UTP) cabling. Often, that switch is on a fiber optic backbone, connecting it to a switch in the main computer center.
Fiber to the desk using a media converter at each connected PC offers an alternative solution, but the media converter costs extra. As long as each PC has a UTP connection, the resistance to converting it to fiber has been a deterrent.
Cliff Walker, who designed the giant local area network (LAN) at the Dubai International Airport, made a presentation on the network architecture he created for the airport that resulted in large cost savings. He called his architecture fiber to the office (FTTO).
The FTTO solution was basically a modification of the traditional fiber backbone with copper-to-the-desktop design. However, instead of putting switches with large numbers of ports in the telecom rooms and running permanent UTP copper to an outlet near the desktop where the PC connects with a patchcord, FTTO runs fiber to the same location as the office outlet where it connects to an inexpensive switch with only several ports—usually four—where PCs can connect with regular patchcords.
FTTO creates the same LAN connection without needing the telecom rooms or the dedicated UTP links for each user. It eliminates the need for racks of patch panels, large cable trays, patchcords and power and alternating current in every telecom room. Each switch is powered at the office using either power cubes or hardwired power.
FTTO works equally well with multi-mode and single-mode fiber. The computer room switches are directly connected to each of the office switches with gigabit Ethernet, shared by four users, providing more throughput for each user than typical networks that use switches in the telecom room with many more ports, a large benefit for wireless access points.
Based on the Dubai Airport network, FTTO saves anywhere from 15-–50 percent in component costs. Obviously, it involves less labor, too, since all the horizontal cabling is gone, but that’s replaced by higher profit fiber cabling installation.
When I describe FTTO to some cabling people, they immediately ask about power over Ethernet (POE). Well, I don’t know any desktop computers that can operate off the limited power available with POE. While there are standards being developed to up POE to sufficient levels to allow laptops to use it, the voltage drop on those tiny wires in UTP cables makes it highly inefficient, if it works at all. If you need POE, for voice over Internet protocol (VoIP) phones or wireless access points for example, you can get the FTTO office switches with POE ports, too.
Passive-optical LANS (POLs)—a variation of FTTO based on fiber to the home—is being used in large networks, especially where security provided by POLs encryption is important. However, FTTO scales better for smaller networks.
So, if FTTO is better and cheaper, why don’t we see more networks using it? In my experience, IT managers don’t want to deal with cabling; they just keep doing the same thing they have been doing for decades. Cabling companies don’t push concepts such as FTTO because they don’t sell as much cable, and few network equipment suppliers offer appropriate products even though the concept of FTTO has been around for decades. Alcatel and 3Com offered products like this in the late 1990s.
Educating the customer helps, but the real way to sell FTTO is the method Walker used at Dubai Airport. Create two basic designs, and compare the costs. Especially in large buildings, such as airports or government facilities, the cost savings should be substantial, and the customer has no alternative to using fiber because of the distances involved. Couple those facts with the virtually infinite bandwidth of single-mode fiber compared to the expected short lifecycle of category-rated UTP cable, the lower initial cost and lower lifetime cost of fiber, and it becomes an easy sell.
HAYES is a VDV writer and educator and the president of The Fiber Optic Association. Find him at www.JimHayes.com.