The Internet continues to drive demand for multiple voice lines and Internet access methods that range from 56 kb/s dial-up modems to DSL. Local exchange carriers nationwide continually face the challenge of limited outside plant infrastructure to serve the increasing number of subscribers demanding multiple voice lines and high-speed Internet access.

In high-density urban areas targeted for asymmetrical DSL (ADSL) deployment, there is often a lack of available physical copper pairs over which to deliver ADSL services. Furthermore, competition from cable operators and competitive LECs threatens to take not only the Internet access business, but also incumbents' voice line revenues.

Hundreds of billions of dollars have been invested in the copper access infrastructure, and more than 800 million access lines exist worldwide. But continuing to bury new cable is increasingly inefficient, given the development of emerging optical fiber technologies. At the same time, replacing the copper infrastructure entirely with optical fiber isn't economical, nor can it be accomplished in a reasonable amount of time. In an era of fierce competition, carriers need interim solutions to keep customers from looking for alternative service providers.

During the same time that DSL transmission technologies were in development, another important step forward for the imbedded copper plant also was being perfected: line powering (Figure 1). Line-powered digital transmission systems extend the capabilities of the existing copper plant. Such systems take advantage of metallic wire's electrical properties and use the wire itself as a powering device for remote units placed in the OSP. These remote units enable new applications that increase the economics of copper to deliver data and voice services to the mass market.

Better copper usage

In a telco central office (CO), existing equipment provides ordinary, non line-powered voice and data services. For voice POTS, dial-up modem access or ISDN, this platform would be a voice switch. But in the case of ADSL service, this platform would be a DSL access multiplexer (DSLAM). This equipment often is referred to as a local access platform. In the absence of line-powered technology, this platform would connect directly to copper pairs that would run out to the customers, with one pair per service delivered.

Alternatively, service providers can use a line-powered system. The line-powered system includes a CO terminal unit, remote terminal unit and the repeaters. The CO terminal unit takes services from the local access platform, combines multiple services, adds a voltage to power the line and sends the combination out of the CO in a digital format. The repeaters recover this digital format and recreate a "clean" digital signal for longer distance transmission. One or more repeaters can be used, depending on the distance from CO to customers.

It is important to note that no repeaters are needed; the CO terminal unit could communicate directly with the remote terminal unit. At the remote terminal unit, line powering is removed from the pair, and the original multiple services that were fed to the CO terminal unit are recreated and sent to the customer premises over "drop" wire connections. In Figure 1, eight POTS services are being sent to eight homes; however, this need not be a 1-to-1 relationship. In a neighborhood where many homes have multiple phone lines, eight POTS circuits might serve three or four homes.

A line-powered digital pair gain system enables more services to be deployed in areas where there is a shortage of free copper pairs. In this example, the system has an 8-to-1 pair gain advantage - what used to require eight separate pairs to deploy now takes one. Seven pairs have been saved, or conversely, the provider has avoided deploying seven new pairs. Because the burying of new cable is expensive, the economics of pair gain systems can be extremely compelling. With a pair gain solution, carriers can extend the usefulness of the existing copper infrastructure, avoiding the expensive installation of new copper (or new fiber).

Figure 2 illustrates the economics of deploying new copper vs. using a line-powered system on existing copper. This analysis shows that for any users more than 5000 feet from the CO, line-powered solutions are more economical than installing new copper. Because of logistical difficulties in installing new wiring, such as delays in obtaining the necessary permits, a line-powered pair gain system often will be the best choice, even when distances are less than 5000 feet.

Carriers should note that such an approach is not possible with optical networks because fiber is not an electrical conductor and requires separate powering for all remote units.

More DSL to the masses

Because a line-powered system uses digital transmission between the CO terminal unit and remote terminal unit, the quality of the signals it carries is preserved. This solves some important problems with data service. For example, residential computer users who have dial-up modems designed to yield 28 or 56 kb/s speeds often find their true connection speeds are less than 20 kb/s. This is because non-digital POTS transmission degrades with distance from the CO, as the signal attenuates and is distorted. Rural modem users most often experience this effect because they tend to have long copper connections to the local CO. However, rural users served by a line-powered digital system would achieve the same modem speeds as if they lived just down the street from the CO.

Another example of this benefit is with DSL data services. Suppose the local access platform in Figure 1 is a DSLAM that serves ADSL or ISDN DSL to users. It is a well-known shortcoming of DSL technologies that their reach is limited; perhaps only 60% of the residences served from the CO are close enough to receive ADSL service.

Many telco customers who order the service are denied because they are "too far away." A line-powered system using repeaters can increase the reach of DSL from approximately 60% to more than 98% because the digital nature of the system preserves the integrity of the high-speed data signaling. Furthermore, because the system is line-powered, the remote terminal unit doesn't require a powering source and battery backup. A system such as the one shown in Figure 1 represents a cost-effective solution to the digital divide between the DSL haves and have-nots.

Safety first

A natural concern with line-powered systems - whose voltages can exceed +/-150 V - is safety. Telcordia Technologies has examined safety issues involving transmitting voltages on the same copper pairs used for signaling and has classified three ranges of line powering that are safe for use. These classifications are A1, A2 and A3, and they describe the voltage and current limits that are safe for barehanded contact. In addition to voltage and current limits, ground-fault interrupt circuits can be used in line-powered systems for more safety.

Although contact with line-powered copper pairs is considered safe for telco operating personnel, line-powered systems do not expose the general public to such voltages. In a line-powered system, power is applied to the pair at the telco office and runs down the pair, either on telephone poles or underground. The pair may travel through one or more repeaters and eventually terminates on a remote terminal unit.

This remote terminal unit sends ordinary low-voltage signals on the copper pairs ("drops") that connect it with the end user's homes or businesses. The electrical characteristics of these drop wires coming into the customer premises are the same as if the customer were served directly from the telco office without the use of a line-powered system. High voltages are not brought into the home, and there is no exposure to the public - only to telco personnel who are trained in electrical safety.

Line powering vs. fiber-fed DLCs

As previously discussed, a line-powered digital system is more economical than new copper. A comparison to installing new fiber also is necessary, but it needs to be done in a slightly different way. Fiber has virtually unlimited bandwidth, and from a fiber node, thousands of services could be generated. Given adequate demand for these services, it seems that fiber installation would make sense. However, in areas where only a few more services are needed, a line-powered system using the existing copper will be less costly. Thus, the analysis needs to be done as a function of the size of the addressable market, not as a function of distance from the CO.

Figure 3 compares the costs of installing a fiber-fed digital loop carrier (DLC) vs. using line-powered copper-fed remote terminal units. The DLC has a high fixed cost: A concrete pad needs to be poured, an environmentally sealed enclosure needs to be built, local power needs to be supplied and a battery backup needs to be put in place. The one-time capital cost of these items is estimated at $100,000. After this point, creating more services such as POTS or DSL data at the DLC involves installing new cards; these are estimated at $150 per service. This assumes that the fiber already is in place; if not, approximately another $5 per foot needs to be added for fiber installation.

On the other hand, an eight-line copper-fed remote terminal unit costs around $500 per service, including installation. There are no large one-time costs, and the $500 per service figure remains roughly constant, regardless of the size of the population being served. The analysis assumes that the node, either a DLC or a line-powered remote terminal unit, is 15,000 feet from the CO.

This analysis shows that in areas where less than 250 new services are needed, using line-powered copper systems is more economical than a switch to fiber-fed platforms. The exact break-even point depends on whether fiber already is available to the node - if so, DLCs become a viable solution at 300 services; if not, the cross-over point moves out to around 500 services.

With a 10-year planning horizon and projected growth rates of about 3% to 4% per year of telephony services - in the form of second lines plus DSL services - it would take a population of more than 800 households to generate a demand for 400 new services. Assuming 2.5 residents per household, this equals 2000 people. There are certainly many areas of the U.S. where 2000 people are within reach of a DLC, but in most places, this is not the case. With a DLC serving roughly a 10,000-foot radius, this works out to a population density of about 180 people per square mile. Most of the counties in the U.S. have population densities below 100 people per square mile.

The U.S. Bureau of the Census reported in 1990 that 80% of the U.S. population lived in metropolitan statistical areas (MSAs) and that these MSAs covered less than 20% of the country's area. In these places, installation of fiber DLC systems may be the correct long-term goal.

However, in the remaining 80% of the U.S. land mass, copper-fed line-powered devices appear to be the correct long-run economic solution. Furthermore, even in metropolitan areas that eventually will see fiber use, copper-powered systems also will play a role because they can be rapidly deployed. Waiting for widespread DLC installation is a dangerous game for telcos in an era of competition; interim solutions are needed to keep customers from looking for alternative service providers.

Now and later

Line-powered digital systems soon will alleviate most of the limitations of the copper access infrastructure: shortages of available copper pairs, limited data transmission speeds and limited reach of DSL services. By solving these problems in a cost-effective way, such systems are giving new life to the copper plant and taking it into the age of mixed data and voice communications. Digital systems boost the throughput of traditional dial-up modems, yielding faster Internet connections. Line-powered systems offer a complement to traditional DSL technologies; they free up pairs for DSL use, as well as extend the reach of DSL to those beyond 18 kilofeet.

Longer term, some overbuilding of the copper network with fiber will continue in areas where population densities are high enough to justify fiber's high installation costs. However, because of the large capital requirements, this likely will progress slowly and probably will not go beyond urban and suburban areas. In less dense areas, line-powered copper delivery mechanisms likely will prevail as the most economic solution to providing data and telephony services.