PDF of this tutorialIn an ideal world, the network provider could start from a clean infrastructure slate and would deploy fiber as deep in the network as economically possible. This would be the most desirable approach to the rural FSN, as fiber provides the most bandwidth potential of any medium, yielding the most flexible growth path for future system expansion and services. The process for deploying fiber, however, is lengthy and a major investment.
Thus, the rural FSN must, if possible, reuse the existing infrastructure consisting of copper, twisted pair for telephony, and coaxial cable for video services. In some cases, a wireless infrastructure provides either the quickest or most economical method of delivering either telephony or video services.
Table 1 shows the cost of fiber only—the cost of installation or splicing is not included—as a function of distance. The figures shown in Table 1 represent the lower end of fiber costs. Factors affecting fiber costs include whether the fiber is housed in aerial or buried-type of cable. Rural areas can easily exceed 5 and 10 miles, so clearly there is a huge economic benefit to equipment that can operate using fewer fibers.
| Distance (miles) |
6 Fiber | 12 Fiber | 24 Fiber | 48 Fiber |
|---|---|---|---|---|
| 2 | $5,808 | $8,976 | $12,672 | $16,368 |
| 5 | $14,520 | $22,240 | $31,680 | $40,920 |
| 10 | $29,040 | $44,880 | $63,360 | $81,840 |
Table 1. Material Cost of Fiber for Various Sheath Counts versus Distance
For urban or small-town applications, the cost of fiber is less as a result of the shorter distances between the CO and subscribers and more easily amortized because of the larger number of subscribers. On a per-line basis, the cost of fiber is not the dominant part of the network. As a result, products built for that market tend to be fiber-rich, and the network is often overbuilt.
What is the cost of overbuild? In the copper world, because every telephone needs a twisted-pair termination and because planners do not know where the next new service is going to be needed, the tendency has been to put in two or even three times the amount of distribution copper needed. The result is a tree-and-branch network with tapered cables (see Figure 4). As the cost per pair for copper is not that high, it was not a really big cost.
Figure 4. Tree-and-Branch Fiber Distribution
A new fiber network tends to look just like a copper network. Instead of every telephone needing a pair, every optical network unit needs a fiber pair. As planners still do not know where they are going to need an ONU, the result is that the cables are still tapered, and the network is still overbuilt. Now, however, the cost for the overbuild is much higher. In fact, the situation is worse, because planners must accommodate both unexpected growth in subscribers and unexpected bandwidth demands from their existing subscribers.
