Copper: Where the action is

Picture this: You're a carrier, and you have customers who would like to spend more money with you. You can serve them with a reasonably small investment--which comes back to you in revenue in a few months anyway--in technology that fits easily into your existing infrastructure. What do you do?

It's not a trick question, and it's not even hypothetical. Of course, you seize the opportunity, especially in the face of the bleak economic situation in this industry today. How you seize it is to take advantage of multi-pair copper technologies, which now allow service providers to deliver fiber-level bandwidth to local loop customers.

These customers--many of them small businesses, others the remote and branch offices of larger corporations--are currently squeezed into DS-1 or DSL bandwidth, due to the lack of fiber in the vast majority of the local loops in North America. Using multiple pairs of copper enables service providers to offer more than enough bandwidth, for instance, to run Ethernet at 10 Mb/s at a distance of more than 12 kilofeet from the central office (see Figure 1).

Depending on distance and the number of pairs used, service providers can boost their customers to multi-megabit bandwidth to handle their range of needs. Those businesses that may not need the 10 Mb/s to extend Ethernet over the WAN could get 4 or 5 Mb/s for improved Internet access or to more comfortably move large data or video files around.

The demand for bandwidth is driven mainly by small to medium-sized organizations, those with 30 to 300 employees, whether they are independent companies or satellite offices of larger companies. Many are found in Tier 2 or Tier 3 cities, where little if any fiber is found in the local loop.

Today, only about one in 10 U.S. business locations is served by fiber, according to a pair of reports from analyst organizations Communications Industry Researchers and RHK. Most of those fiber-served locations are in specific areas of Tier 1 cities. But nearly all of those businesses are served with copper. In fact, you'll find a lot of analysts agreeing that copper is where the action is and will be. As carriers pursue the revenue that they can get from sites already on their networks--rather than worrying about extending their networks to additional sites--expanded capacity is a vital tool.

The bandwidth-hungry businesses may want and need more, but their needs (and budget) still fall far short of the capacity that a DS-3 would deliver to them. For them, the ability to get smaller jumps in bandwidth will suffice for now, whether that bandwidth is in the 4 to 5 Mb/s range or up to and beyond 10 Mb/s.

There are other alternatives available to serve last-mile customers, but for established service providers, increasing the capacity of the existing copper infrastructure makes the most sense. When you measure the use of existing copper against options such as new fiber installation, free space optics, satellite or broadband wireless, copper is the low-cost choice. This is especially true when you factor in installation and maintenance costs and the skills and training needed by installers.

Three are three main approaches to boosting bandwidth with copper. Each of those approaches has certain strengths, depending on the carrier's particular needs. They all have two things in common. First, they use multiple spare copper pairs and multiplex data across those pairs to achieve the greater reach and capacity, and second, they are based on the G.shdsl standard. G.shdsl offers superior spectrum compatibility and high rates for symmetric transmission over copper. 

The three main approaches are inverse multiplexing over ATM (IMA), multi-link point-to-point protocol and multi-link frame relay (MLPPP/MLFR), and multi-loop DSL (MLDSL).

Comparison of copper-based high bandwidth solutions
Solution	Bandwidth efficiency	TDM support	Capital investment	POTS support	Latency
IMA	Low	Yes	High	No	Low to medium
MLPPP	Low	No	Low	No	High, unpredictable. OK for data, not for TDM
MLDSL	High	Yes	Low	Yes	Low

IMA and MLFR/MLPPP are typically applied to DS-1 lines. MLFR is used to provide higher-bandwidth fractional DS-3 frame relay services, while MLPPP is often applied to DS-1s and ISDN B channels to provide higher-speed Internet access and transparent LAN services. MLDSL is applied directly to copper loops.  

IMA is typically used to bond DS-1s together to provide ATM service to subscribers who are not served by fiber. It is based on a 6-year-old ATM Forum standard that defined a new user-to-network interface, which specifies how a stream of ATM cells are spread across multiple DS1 circuits. The UNI rides on top of the physical interface and uses IMA Control Protocol to perform the inverse multiplexing, spreading out the data cell by cell.

IMA can be used to deliver high-speed Internet access, but with a few drawbacks. First, IMA is based on ATM equipment, which is more expensive than an Ethernet alternative, for instance. Second, the IMA protocol and encapsulation for IP data do mean a loss of throughput to the ATM overhead. As much as 20 percent of the available throughput may be given over to the overhead. The ATM IMA protocol also requires all ports to operate at the same data rate, which works well when bonding DS1s together, but if applied directly at the modem, it forces all the copper pairs' transmission rate down to the lowest common denominator.

If all copper pairs being utilized are clean and operating at their optimal data rate, IMA can operate effectively. But if one pair out of say, eight, is degraded due to a bridge tap or other reason, then all pairs will transmit only at the maximum data rate achievable by the degraded pair.

Multi-link protocols offer an approach that does not involve ATM and avoids the consumption of available bandwidth by the cell overhead. Both multi-link protocols, MLPPP and MLFR, group together several DS-1s, or in some cases E-1s, which may be transported over copper with HDSL or any other transport technology.

MLPPP is typically deployed in the ISDN market, where it is an effective means of bonding B channels to provide 128-kilobit data transport. In the Internet access market, it can bond DS-1s for higher-speed Internet access or transparent LAN services.

MLFR is utilized to bond DS-1s to provide fractional DS-3 frame relay service. Both boast low transmission overhead and enable automatic link recovery and bandwidth redundancy. On the negative side, both of these protocols use relatively large packet sizes, which introduce latency into DS-0 or DS-1 TDM transport applications.

Multi-loop DSL, or MLDSL, differs from the other technologies in that it does not base its per-pair transmission rate on DS-1. It has the ability to transmit at a wide variety of rates, all the way to 2 Mb/s and above. The inverse multiplexing technique that MLDSL employs, as it operates between Layers 1 and 2, is designed to reduce latency and overhead and allow greater throughput.

It can bond copper pairs at different line rates, with the ability to transmit data on each pair according to its maximum capability. This can be an important point in those cases where a pair may be degraded, because the overall effect on throughput is lessened.

MLDSL can operate even if a pair in the bundle goes out of operation, and if managed properly, can maintain the total required bandwidth even without the defective pair. Conversely, if an additional copper pair is added to the bundle, it is immediately utilized to increase bandwidth. Using this approach, MLDSL has the capability to offer loop failure protection.

One other advantage of MLDSL is that it is not limited to operation over spare copper pairs. In situations where there may not be enough available spare pairs, MLDSL can also accomplish its bandwidth-aggregating tasks by sharing the frequency of a POTS line. This is accomplished through the use of a passive splitter, like ADSL.

Work on MLDSL has intensified in the standards bodies as interest has grown in the technology.

There are differences among the three technologies in the way they accomplish their multiplexing. IMA and MLPPP/MLFR are typically implemented at the edge switch, being fed by DS-1s, with each port operating at a DS-1 rate. MLDSL performs its multiplexing at the copper interface, which allows a wider range of data rates.

What about the services that carriers can extend, or launch, once they have more copper-based bandwidth at their disposal?

Some obvious answers to that question can be found in the Ethernet arena, where a multi-pair copper solution can provide the 10 Mb/s needed for native-rate networking. Carriers could offer services such as:

• Local area network extension--This service would connect all of a customer's regional offices so that all would appear to be on the same LAN. The customer payoff is that everyone can expect similar performance and access to all network resources

• Intranet or Extranet virtual private network--This service would link a customer's remote intranet sites or make possible links to their suppliers, business partners, or customers

• Internet access--This service would create a dedicated connection for the customer to the local point of presence of an ISP.

There is a strong argument for the rollout of Ethernet services, both in gauging what customers want and the economic scenario for carriers. Infrastructure equipment costs for Ethernet are lower than for frame relay or ATM, and nearly all networking equipment and hosts connect via Ethernet. Using an Ethernet service for interconnections would simplify network operations, administration, management, and provisioning.

In establishing their own business case for high-bandwidth copper solutions, there are a number of important issues that service providers should consider. For instance, they need to weigh the up-front costs and analyze how long it will take to achieve a return on investment. Cash flow, in this time of limited shareholder confidence, has become much more important than the long-term investment. Multi-pair copper equipment should be able to generate income as soon as installation is complete, and though the solution may be scalable, it should be cost-effective beginning with the first subscriber. The total return on investment period should be no longer than a few months, even when all maintenance, administration, equipment and other costs are considered.

The solution should yield a significant bandwidth increase over what is currently in place, and should be easy to install and maintain. Tinkering with the service provider's network can be risky, and no solution should be disruptive to current operations. By the same token, the solution's management tools and guaranteed levels of service are vital issues in terms of the network infrastructure.

Other issues to consider are standards compatibility, to guard against any interoperability issues and to help assure that the equipment itself won't quickly become outdated, and robustness, the ability of the solution to deal with the varying qualities of existing copper pairs and the provider's need for high availability and reliability.

Service providers are clearly in need of a practical, cost-effective way to bridge a gap, between the bandwidth they can offer now in the fiber-poor local loop and what their customers want. Multiple-pair copper technologies may provide the solution until the day when fiber is as plentiful in the local loop as it is in the core network. The equipment takes advantage of the existing infrastructure, keeping capital costs down, and allows providers to begin generating revenue immediately.

If providers choose the equipment wisely and manage the new services well, they can shrink the return on investment period to as little as a couple of months. It makes a business case that is surely worth investigating.

Randy Nash is vice president of business development for Spediant Systems, Red Bank, New Jersey. He can be contacted at randyn@spediant.com.

Visit Spediant Systems online.