For the first 100 years, the outside plant (OSP) was made mostly of copper, plastic, wood, and concrete. Telephones were connected to copper wires, called local loops, and a telephone call was made by establishing an electrical circuit between the telephone set and the Central Office (CO).

The large feeder cables nearest the COs were called underground cables, although not all were under the ground. Underground cables were connected to smaller aerial cables, not necessarily above the ground, which passed every home and business. A telco’s customers were called subscribers, because that’s how they arranged for telephone service.

The OSP was designed, built, and maintained to assure the integrity of electrical circuits from COs to subscribers in the surrounding areas, called exchanges. The electricity was distributed from the CO, and it was the responsibility of the inside plant guys at the CO to assure that there was always enough of it, and that there were batteries and generators to keep it coming, even during an AC power outage.

That 70’s…and 80’s…and 90’s Show
Digital Loop Carriers (DLC) were introduced in the 1970s as substitutes for some of the copper in feeder cables. A DLC consisted of a Central Office Terminal (COT) and a Remote Digital Terminal (RDT) connected together by one or more high-speed digital links. Although a digital T1 line required 2 copper pairs, it had the potential to carry 24 telephone lines. Early DLCs were called pair-gain systems because they could deliver 24 subscriber lines on each T1 link.

Integrated Digital Loop Carriers came along a little later. They allowed RDTs to be connected directly to digital switches, eliminating the COTs and saving a lot of CO line equipment. IDLCs also supported concentration of more subscriber lines on the same number of T1 links. This allowed a single RDT to serve even more subscribers.

Since a DLC interrupts the electrical circuits between subscribers’ telephones and the CO, it must be equipped with a power plant that is, essentially, a miniature version of the power model from the CO. Instead, the RDT gets its power from an AC service ordered from the local power utility. A rectifier converts the AC power to DC for the DLC equipment and charges the batteries that operate the site during a power outage.

Telcos soon learned that a DLC had to be protected from the elements by a tough environmentally-sealed cabinet. Cabinets designed to keep the elements out also keep heat generated by the equipment in; thus fans and heat exchangers were often required. All this additional equipment added to the power consumed at the remote site, but it didn’t seem like a big deal. The monthly electric bills went up a little, that’s all.

Optical fiber came to the OSP in the early 1980s. Telcos learned that fiber feeder cables were much smaller than copper cables, and they carried more T1 links. So, they stopped building copper feeder cables. Telcos also realized that fiber feeders made it possible to deploy larger DLCs serving hundreds of customers. Larger DLCs required more power and even more secure environmental cabinets with more elaborate cooling capabilities. But, telcos thought they could deal with a few more dollars on the monthly electric bills.

Large DLCs serving hundreds of lines needed automated capabilities such as testing and remote service provisioning, integrated SONET terminals and digital cross-connect functions. The integration of these functions into DLCs, beginning in the early 1990s, gave rise to the Next Generation Digital Loop Carrier (NGDLC). An NGDLC can serve up to 2,000 lines, more than many COs, and consumes a similar amount of power. Still, no big deal. Providers just called the power company and ordered a 240V service. (See Figure 1.)

A Powerful Paradigm Shift
For the first 100 years, the telcos built more OSP almost every year. They had to in order to accommodate the growth in access lines. Access line growth peaked in the last half of the 1990s, and telcos scrambled to build fiber feeder cables and NGDLCs to keep up with demand. By 2000, about 25% of all subscribers were served by some 220,000 DLCs and NGDLCs - or about 10 times the number of COs in the U.S.

Don’t forget that all of these DLCs and NGDLCs are powered from the AC grid, equipped with back-up batteries and generator hook-ups, and protected from the elements in environmental cabinets designed to withstand a hurricane. Suddenly, almost before we noticed, the OSP had 10 times the powered sites as COs, and the telcos had electric bills that were more and more expensive to pay every month.

Interestingly, the Communications Act of 1996 opened the OSP to competition, and access line growth flattened. In 2001 it began to decrease. Today, telcos are losing about 10% of their lines per year, and there’s no end in sight. Now they are scrambling to rebuild their businesses on a new technology foundation: IP. Telcos must find ways to rebuild the OSP of yesteryear with the technology of the Internet using IP packets and broadband links.

One way to rebuild it is to replace it all with fiber. But, that might take awhile. Even after a quarter century of building fiber, the OSP is still 85% copper. Even so, some telcos are investing billions to build fiber all the way to customers’ homes (fiber-to-the-premises or FTTP).

One supposed advantage of FTTP is that the OSP is completely passive, with no cabinets full of electronic equipment between the COs and customers. Of course, there’s an Optical Network Terminal (ONT) at every home, plugged into an electrical outlet. The difference is that the customer pays for the power, not the telco. Interestingly, ONTs actually consume 3 or 4 times as much power as CO or DLC line cards, and telcos are deploying millions of them. So, FTTP systems add millions of powered remote sites in the OSP, and will consume megawatts more power than copper or DLCs.

Another way to build broadband OSP is to install broadband remotes relatively close to subscribers in order to use existing copper distribution cables and Digital Subscriber Lines for the final few thousand feet. This is called FTTN: fiber-to-the-node. FTTN is supposed to cost less than FTTP, because there are fewer new fiber cables to place and there are no ONTs. But, the nodes of FTTN must be within about 2,500 feet of the customers they serve, much closer than an NGDLC. They are chock-full of DSLAM equipment that consumes a lot of power and dissipates a lot of heat. They must be protected by environmental cabinets, cooled by heat exchangers, and backed-up by batteries. They must be plugged into the AC grid. And the telcos must pay more electric bills…every month…forever.

Most FTTN sites are at FDIs (feeder-distribution interface), locations, where the feeder cables meet the distribution cables in the neighborhoods. In carrier serving areas (CSAs), served by DLCs or NGDLCs, there are often several new FTTN sites, one at the DLC and 3 or 4 more at FDI locations inside the CSA. Sometimes the area served from an FDI is too large; i.e., the copper loops are too long, and have to be broken up into smaller distribution areas, each with its own FTTN site. All these new FTTN sites require AC power, much more per line than DLCs and NGDLCs. (See Figure 2.)

So, a telco that deploys FTTP requires its subscribers to pick up the tab for the power to run the ONTs at the far ends of its OSP. And a telco that adopts FTTN installs hundreds or thousands of new remote sites in the OSP, all connected to the AC grid. Either way, tomorrow’s OSP winds up consuming more power than COs ever did, and will require a lot of hands-on attention to keep it running.

The irony is that neither telco can retire existing DLCs and cabinets, nor do much about the power they consume. They must keep the old circuit equipment in service until they either lose all the old access lines to competition or move them over onto the new broadband OSP. This will take years, perhaps decades. In the meantime, gigawatts of power will be consumed and billions in electric bills will be paid.

Another irony is that the OSP goes from being the most stable and reliable part of a telco’s network to being the primary consumer of energy and the most fragile in terms of remotely deployed electronics elements. It consumes a lot of gasoline, too, because almost anything that needs doing in the broadband OSP requires someone to get into a truck and drive to a remote location to do it. Telcos have spent decades automating their COs and long haul networks, but the surfaces of the OSP remained almost unscratched by automation. (See Figure 3.)

The ultimate irony is that, just when the world is becoming aware of the costs of electricity and gasoline and their impact on the environment and the future, telcos are building broadband OSP that will consume more of both.

Powering in the Right Direction
Maybe we should do something about these challenges to power the OSP. Here are six ways we can actively begin to change our OSP power paradigm.

Option 1: Accelerate the deployment of packet technology. IP is already widely used in telcos' backbone networks. IP enables more efficient sharing of a physical network by the information and services traversing it. That means less network; less equipment; fewer people to operate and maintain equipment; less space, power, and air conditioning, etc., etc., etc.

Softswitches are already switching billions of minutes of long-distance traffic, and are beginning to be deployed to replace conventional circuit switches in COs. Let's build more IP-based OSP: FTTP and FTTN, as quickly as we can, to deploy IP technology into the OSP on a broad basis.

Option 2: Be aggressive about moving existing customers and services onto the new IP OSP.

Option 3: Start unplugging the old OSP. DLCs were always engineered for years of growth, so they were seldom fully-loaded with customers. We've seen several years of negative growth, so there's a lot of OSP equipment sitting out there in cabinets using power, taking up space, and dissipating heat. At least half the DLC and NGDLC lines installed in environmental cabinets are not being used, and will never be used again. Unplug 'em. Get 'em out of the cabinets. Reuse the old cabinets for new IP OSP.

Option 4: Be sure that the new IP OSP equipment is as energy-frugal and efficient as possible. The big telcos have recently set new standards for equipment power consumption, and put metrics in place to measure and compare it. Verizon, for one, is requiring that all new equipment use 20% less power than earlier generations. These initiatives will have very little impact on current power consumption. But, they're a good step for the future.

Option 5: Get serious about automating the new OSP. Why build FTTP equipped with physical cross-connect points in the OSP? It guarantees a visit to the remote site every time a new subscriber is connected or disconnected, or when an optical connection is opened up for testing. There are new, automated optical cross-connect systems coming out that will keep the hands out of the plant and the trucks off the roads.*

Why build a new FTTN site right next to an existing FDI cabinet and continue to dispatch technicians to run jumpers to connect or disconnect broadband subscribers? There are new automated systems available that enable telcos to take control of their copper assets and remotely manage their OSP. **

Option 6: Telcos must look at large-scale deployments of solar power systems. It's estimated that telcos consume as much as 2% of all electricity that is generated. Telcos consume most of their energy at relatively few sites: COs and DLC remote sites. A lot of the power consumed by telcos is DC power, just the kind produced by the photo-voltaic panels in solar power systems. The sun rises on every CO, and on every DLC remote site, and on every FTTN site - every day. Why not install solar systems on most of them, wherever practical? Since a majority of the power consumed is in the OSP, why shouldn't OSP take the lead?

This isn't a new idea. Bell Labs invented the photo-voltaic solar panel in the early 1950s, and installed a solar power system to power a small remote site in Southern Bell in 1955. Twenty years ago, OSP® Magazine published a story about a solar-powered DLC installed at a remote site in the High Sierras where no AC power was available. (See "Sun Power For High Sierra's Telephones", OUTSIDE PLANT® Magazine, March 1989.) Solar power systems are installed on millions of homes and thousands of businesses all across the U.S. There are tax credits and other benefits that cover a substantial part of the first costs of a solar system. The sun always rises on the OSP, so solar power can cut the substantial and growing costs of powering the OSP every day, every month, every year…forever.

It's time for telcos to step up and take the lead in applying solar power. Solar power can cut a telco's energy costs by a significant fraction, even more if we slim down and automate the OSP to align it with the broadband IP future. It's time for telcos to commit to solar power, and to put their RFPs where their mouths are. If they do, suppliers will respond with solar power equipment to cut costs, reduce emissions, and make more reliable networks.

Let's start with the OSP!

Endnotes
* Calient Networks. www.calient.com
** Sagamore Systems. www.sagamoresys.com

Kermit L. Ross is founder and Principal of Millennium Marketing, a consultancy providing research, strategy, and planning services to telecommunications equipment suppliers. He has more than 40 years of experience in the telecom industry. Kermit can be reached at 972.668.3675 or email klross@gradecom.net.

What's your experience with this? Tell your fellow readers now!