TVA Rural Studies

Improving Rural Telecommunications Infrastructure

Bruce L. Egan
Columbia University

6. Network Upgrade Costs

Figure 2 provides an illustration of an advanced digital rural subscriber connection. The basic loop architecture is similar to today's average rural POTS loop except for a few features. Assuming that the basic POTS loop meets the maximum length for high quality digital service e.g. 12-l8 kft.) and that the serving C0 already houses a modern digital switch, the placement of sophisticated electronic equipment located in the three shaded boxes between the subscriber premises and the C0 enables the subscriber to use a range of new digital services.

Upgrading the loop plant of rural telephone subscribers for digital service presents a financial dilemma. A high percentage of existing subscriber loops cannot support an acceptable level of digital transmission, even for existing services. Regular voice telephone service requires much more bandwidth in digital form than in analog form. Most rural loops are engineered to support analog voice at 3–4 kHz, and very low speed data service up to 9.6 kb/s. To attempt more than this risks intolerable errors in transmission. Thus, the motivation to upgrade the rural loop plant is that current band-widths will not support the use of many new service applications.

It would be misleading to conclude from the data on rural company loop investment that the upgrade problem is simply solved over time by replacing investment through rapid depreciation. Increased cash flow from depreciation, an important source of funds for new loop plant, also implies rate increases for current subscribers, increased subsidies from others, or both. In addition, the new loop plant is nominally more expensive than the old, even with technological advances, because of inflation in prices.

Generally, the main problem with upgrading rural subscriber loops to digital service is the presence of loading coils. These must be removed by cutting out the coils and replacing the cable at the load coil point. Alternatively, loop carrier or remote switching terminal equipment may be installed. Normally, this is all that is required in the physical loop digital upgrade. However, some rural telephone companies still have old "non-filled" cable in their loop plant. This may not support high-quality digital service even at low speeds if moisture has penetrated the cable. Nevertheless, analog voice is generally acceptable on non-filled cable. The financial requirements for upgrading "gel-filled" cable rural loops for digital service are not too much of a burden for current telephone company construction budgets over a reasonable period of time. However, for "non-filled" cable loops, a costly and aggressive rehabilitation program may require external financing. The process of replacement will speed up since the remaining non-depreciated useful life of non-filled cable is relatively short (it was last installed in the early 1970s).

6.1 Narrowband Digital Service

Chart 6 presents the base case costs for current narrowband rural LEC loops. The estimated cost of upgrading existing rural loops to provide for ISDN service is only about $100 to $200 per subscriber (again assuming that the loop is qualified in terms of length and electronics) For non qualified loops (featuring load coils, non-filled cable, etc.), the average cost can be anywhere from $50–$2,000 per subscriber²³. This only represents the average; some customer loops will be even more expensive to upgrade, such as where spatial distribution of subscribers was not conducive to sharing facilities. One goal of the upgrade, de-loading rural loops, could be very expensive when there is no cost justified possibility for shortening the dedicated portion of the subscriber loop through the use of a ISDN compatible remote subscriber terminal (RST) or digital loop carrier (DLC) system. The state-of-the-art loop architecture assumes that a fiber trunk connects an RST to a digital host CO (see Figure 2).

6.2 Broadband Digital Service

Based on a broad based analysis of existing (1992) RUS company cost structures, the monthly cost of deploying a rural broadband network is estimated to be between $92–$132 a month per line depending upon the period for deployment (10-20 years)²⁴.

Whereas rural network upgrades for narrowband digital service are based on maximum loop lengths of 18 kft. from the switching node, higher bandwidth and power requirements of switched broadband networks will require a smaller serving area featuring loops of only 6–12 kft. depending on the services contemplated and the specific network design. This raises costs considerably. For example, reducing a maximum serving area distance from 18 kft. to 6 kft. means that 9 network nodes are required vs. only one.

The digital loop diagram in Figure 2 indicates where electronics may be installed to allow subscribers to upgrade service for broadband capability like entertainment video service. Recalling that the downtown area of the rural exchange might well be within the 6 kft. limit, this situation certainly favors that area over the outlying rural area in any upgrade decision. Chart 7 provides an estimate of a rural LEC’s broadband loop upgrade using Hybrid Fiber Coax (HFC). Assuming that the maximum number of households served per HFC network node is 480, Chart 7 shows how per subscriber costs might be expected to vary as subscriber density varies (i.e., as one moves out from the downtown area toward the rural areas of the exchange). Subscriber access connections within the dense downtown area may be upgraded to broadband service for $1,000, while serving subscribers in the outlying rural areas of the exchange can cost up to $10,000. The illustrative costs in Chart 7 are for subscriber connections, and do not include the costs of upgrading other network and non-network functions including sophisticated broadband network system hardware and software and programming service. One estimate is that this could add another $400–$1,500 per subscriber.

Another possibility for providing broadband telecommunications to rural areas is through upgrading the existing rural coaxial cable systems with fiber optic trunk lines and interconnecting to the public switched telephone network. For a truly integrated broadband system, this usually generates per subscriber costs similar to those already discussed for telco network upgrades. There are other (even more sophisticated) methods of providing broadband services to the home, but the costs of these alternatives are generally equal to or higher than the HFC network upgrade.²⁵ Fiber to the Home systems are touted as being the ultimate in broadband telephony featuring high quality bandwidth on demand with capacity for any conceivable service. The costs of such systems for rural applications are currently so high as to not even be seriously considered by rural LECs, however, this conclusion in no way detracts from the great potential of fiber optic trunk network systems in rural settings. ²⁶

6.3 Wireless Alternatives

For situations where it is simply too expensive to use the recommended loop architecture, there are several alternative choices including satellite, point or multipoint radio, and cellular radio. These alternatives must be evaluated on a case-by-case basis, including an estimation of the cost of an efficient connection to the public wireline network.²⁷

Digital wireless technology could potentially become a cost effective replacement for fixed wired telephone service for everything from POTS to broadband service. In particular, new digital wireless cable networks are already competing with traditional wired cable in urban areas and it is widely believed that the new digital cellular Personal Communication Networks (PCNs) will provide a cost effective alternative to both the fixed and mobile cellular telephone systems in service today, all at competitive prices.²⁸ However, rural areas pose a special problem for successful deployment of cellular systems for fixed telephone service. Furthermore, while PCN systems using small cells (microcells) are optimal for low power operation in urban settings (i.e., dense market areas), they are not cost effective in rural areas because of the sparse number of subscribers who can share a single base station in a small coverage area.

Due to the distances involved in a rural setting, the power levels for transceiver base stations needs to be much higher than in urban cellular markets or it is not possible to reach enough subscribers to make the investment worthwhile. Too many low power antennae sites would be required to cover rural areas in a cost effective manner. Current microwave radio systems for rural telecommunications (dubbed BETRS by the FCC) are very expensive to deploy and operate and tend to be cost effective only in the thinnest rural markets, or where terrain will not permit wired subscriber connections.²⁹

Many recent articles have touted the virtues of using wireless access at a cost effective substitute for wired access in rural areas.³⁰ Hatfield, Paulraj and others show that in the thinnest markets (0–100 subscribers per square km), fixed microwave radio (i.e., BETRS) systems may be cost effective to deploy. Other authors show that cellular systems using large cells (macrocells) are also cost effective in many rural markets including downtown areas. Figure 3 provides a broad gauge look at the relative cost effectiveness of macrocell and microcell wireless access systems vs. wired access.

Raw cost efficiencies aside, much of the problem with deploying wireless networks in rural areas lies more with the long head start and continuing inertia of wired service and the ingrained preferences of telephone company managers and engineers for the old (well understood) way of doing things.³¹ A second important problem to overcome is the current federal rules governing the provision of digital cellular service in rural areas and the limited radio spectrum frequency which has been licensed for use by rural radio systems.³² Currently, rural cellular service must be provided under restrictive conditions imposed by the government on radio frequency use. Rural cellular providers must share radio frequencies with existing high power paging services, causing interference problems. Channelization schemes used by current urban area cellular radio licensees are not optimal for use in rural areas. Power restrictions are too low and the radio carrier channels are too narrow to allow for a single channel to be shared cost effectively in a rural setting.

If the government would allocate sufficient dedicated radio frequency spectrum (e.g., 20 MHz) and increase permitted power levels, then cellular equipment manufacturers and network operators could use state of the art digital access techniques such as TDMA and CDMA and wide carrier channels (e.g., spread spectrum). This would allow rural cellular service to become a cost effective replacement for expensive wired POTS access arrangements and could reduce the costs of broadband network upgrades.

For subscribers in rural areas which are truly remote (perhaps even unserved), new digital satellite systems offer the best hope of obtaining high quality digital telephone service. Beginning in l996, many new systems will be launched, providing coverage over the entire continental US. Initially prices for these systems will be very high and some subsidies may be required to make it available. One of the main reasons why satellite service has not been viewed as potentially competitive with wired service is the annoying (and heretofore unavoidable) delay time associated with voice transmission on the system uplink and downlink segments (250 milliseconds). Many new digital satellite systems have overcome this quality differential by using Low Earth Orbit (LEO) satellites which feature only a fraction of the delay time.³³

The recent (and rapid) introduction of digital satellite television using Direct Broadcast Satellite (DBS) service demonstrates that rural areas will be able to benefit substantially. The cost of this technology is not distance sensitive and therefore rural subscribers can finally obtain equivalent service at equal prices in a market setting.³⁴

Jump to Section:
Contents, 1, 2, 3, 4, 5, (6), 7, 8, Tables, Footnotes, Charts & Figures

Back to the 1996 Rural Telecommunications Workshop Homepage