Case by Case: TIA FOLS overhauls LAN cost model

The Telecommunications Industry Association’s Fiber Optics LAN Section (TIA FOLS) recently released its third generation interactive cost model for estimating LAN infrastructure costs. According to FOLS members, instead of simply updating the model’s second generation, the new model was designed from the ground up in response to recent industry trends.

“The previous models lacked a couple of significant parts. One was that they needed to use real world pricing,” says FOLS member Dan Harmon, an applications engineer and consultant programs supervisor at 3M. “The other was that we wanted to stop focusing on a specific fiber technology-in other words, one connector, one type of cable-and really just use parts that are [generally] available to people.”

Significantly, the updated cost model employs aggregate, representative pricing data derived from www.peppm.org, an educational technology and bidding program. In all cases, the group says the pricing used in the model represents an average price taken from at least three different manufacturers; the updated model also expands on previous models by focusing on the support of converged networks and comparing standards-compliant architectures.

“In the last two or three years, the market’s changed quite significantly. Certain components have dropped in price, and new architectures, which have become more utilized, are really changing how people are reviewing their infrastructure options,” explains FOLS spokesman Liz Goldsmith. “We knew that today people are looking at infrastructures to support converged networks, and we wanted to choose the types of switches that would allow that kind of support. We made this cost model a little bit broader in its scope, a little bit more relevant to the people who are making decisions.”

FOLS first released its fiber versus unshielded twisted-pair (UTP) copper cabling cost model in 2002; the interactive model was developed by FOLS and Pearson Technologies. The tool was designed to offer network planners a cost model that examined an entire network infrastructure deployment, including labor costs, as opposed to traditional cost models, which the group says typically considered only electronics, cabling, and connectivity.

“One of the main differences of this model versus previous models is that this one’s very simplified,” observes Rodney Casteel, FOLS member and a fiber-optic sales specialist at CommScope. “Basically, here we just look at four different architectures. In the past, we looked at different scenarios of combinations of components and electronics. With this one, we’re consistent on the electronics and components-what we’re changing is the [network] topology.”

The new model compares the costs of a LAN based on a hierarchical star architecture, constructed from copper UTP horizontal cabling with a fiber riser backbone, to three all-fiber architectures based on centralized (also known as collapsed backbone) cabling as well as low-density and high-density cabling infrastructures based on the recently ratified “fiber to the telecom enclosure” (FTTE) standard.

“In the last model, we did reference the fiber to the zone, which was something people were starting to use, but at that time it was not a standards-compliant architecture,” notes Goldsmith. “FTTE is a standard,” adds Casteel, “and it’s probably the most efficient architecture you can use today, because you’re coming into the enclosure with a 1-Gbit uplink, and then you’re coming off the front of the switch with eight 100-meg ports, so you’ve got 800 megs going back to 1 gig, and it’s a totally nonblocking architecture. I can come off the front of that mini-switch with copper, which is just a small cable run to the work area, and then my moves, adds, and changes become very simple, because it’s just a small cord I’m moving around versus pulling a whole new home run back to a closet somewhere. The switches [in the FTTE architecture] are very-low-cost, so it enables a lot of flexibility in design and in the implementation of convergence. In the model, we do a low-density and a high-density [FTTE] solution. The high-density solution saved us the most money, because we were getting more users per enclosure, and so it was a lower cost overall.”

“Although [the FTTE architecture] is inside of a building, [it] actually mimics FTTH,” notes Harmon. “It’s really an interesting architecture. It’s not all fiber, but it does use copper in a more efficient way. It’s using copper for shorter links, which is what copper’s good for. Equipment in FTTH is really geared toward convergence; it’ll usually have a fiber link coming in the back, and off the front, it’ll not only have copper ports, but also plain old telephone ports or ports that will pass video information.”

According to the group, as with previous generations of the cost model, the results of the comparisons among infrastructures in the latest model depend on users’ unique input. However, the group says that in the latest model, when using the sample scenarios, the installed first costs of fiber-based infrastructures are shown to be either less expensive or within 5% of the installed first costs of the traditional combined fiber backbone and UTP horizontal architecture.

“Many people know that optical fiber offers several benefits over the life of the network, such as reduced maintenance costs, ease of upgrading, and reduced downtime-but the model developed looks specifically at installed first costs,” explains Harman. “Lower lifecycle costs are an important benefit of deploying fiber, but they are also more difficult to quantify. We felt that installed first costs were so critical to end users that we wanted to focus on them in the model. Additionally, the model directly dispels the myth that fiber-based networks are always more expensive than copper networks.”

Matt Vincent is associate editor at Lightwave.