The DSL market has undergone a complete makeover. There are now nine variants of DSL, and it seems that every year another variant is invented. CLECs and ISPs have realized the huge pent-up demand created by the large number of small and medium-sized businesses, the huge expansion of the SOHO market, and the growing needs and requirements of teleworkers and are heavily targeted these markets. During 2000, the U.S. DSL marketplace saw explosive growth, adding more than 1.9 million new subscribers for a growth rate of 382 percent from the end of 1999. This impressive growth of the DSL market shows no signs of slowing in the coming months.

DSL builds on the work of engineers in developing dial-up modems to offer data over voice access to the copper spectrum. DSL uses the existing copper plant to provide data over voice, independent voice and data transmission, and plain old telephone service (POTS). Because DSL is a loop technology, rather than a broadband access technology, it can be deployed in many different configurations and in many different types of telecommunications equipment. In a few years, we will probably see DSL deployed not only on all digital loop carriers (DLCs), but also on channel banks, asynchronous transfer mode (ATM) switches, voice switches, and on next-generation DSL access multiplexers (DSLAMs).

Demand for bandwidth has led to several technological approaches developed to provide broadband access to business and residential customers. DSL is not the only path to achieving high-speed broadband access. Several competing technologies, such as cable modems, broadband satellites, and broadband fixed wireless services, are trying to achieve the same goal—to provide high-speed broadband access to consumers. This report examines the market strategies and deployment of these alternative services. Although DSL has numerous advantages, other means of obtaining high-speed access might be preferable to some customers. Several fixed wireless technologies are beginning to challenge traditional wireline and cable for the data marketplace.

Widespread installation of DSL has been impeded by physical constraints of the local loop (number of load coils and bridge taps as well as its distance from the central office [CO]), inability to access customers served through a DLC, interoperability problems between products, and spectrum compatibility issues. The need for automated provisioning and flow-through of orders, better training and customer service, a shortage of experienced technicians, and the lack of true plug-and-play modems have also hindered the spread of DSL.

Self-installation continues to be the holy grail of the DSL industry. Finding a way to move beyond service-provider CPE channels and into the retail market is the key to a successful mass-market deployment of DSL. DSL needs to become as easy to purchase and install as an analog modem is today. Consumers need to be able to purchase a modem at an electronics or other retail outfit, install it on their PC, and immediately sign up for DSL with little or no support.

If DSL providers and vendors can achieve this goal and if customer service and technical support of the service provider can follow, then a mass-market deployment will happen much sooner than expected. Besides implementing automated provisioning and configuration system and processes, the biggest obstacle facing the industry is the need to significantly improve the customer and technical support provided to all potential and actual subscribers. If these obstacles can be overcome, DSL will have a bright future. Without a significant improvement in customer service, the deployment of DSL to the mass market will fall far short of expectations.

For DSL to be successful, it must not only meet the expectations generated by its marketers, but also guarantee service quality and be as easy to install as an analog modem with a dial-up ISP. Otherwise, the advantages it offers (e.g., high-speed access, low cost, videoconferencing, access to company LANs, and other applications that require higher bandwidth) will be reduced substantially.

Without service guarantees, small and medium-sized businesses are unlikely to use the service for their mission-critical applications. Quality of service (QoS) guarantees are essential to any business and have become one of the main value-added services that DSL service providers can offer to differentiate themselves from other providers. Businesses have always paid more to ISPs that offer tier-1 Internet service, and the same can be said for service providers that offer service-level guarantees to small businesses.

Newer technologies are not only reducing the costs of transforming inputs and outputs that result in goods and services, but also are decreasing interaction costs, i.e., the costs that are incurred in getting different people and companies to work together to exchange goods and services. According to Roger Ferguson, these interaction costs account for about 55 percent of all labor costs, with some industries, such as financial services, having interaction costs as high as 70 percent of labor costs.

The industry needs to focus on measures to strengthen customer service initiatives, review current support processes and service levels, and map future demand for customer support. Otherwise, deployments of broadband access infrastructure will be significantly impaired. Companies need to ensure that both their customer and technical support teams, as well as their partners or affiliates, are well trained and responsive to customer needs. Otherwise, these companies risk permanently endangering both their business success and the rollout of a broadband infrastructure in the United States. We see this total inattention to customer service as the biggest danger facing all broadband access and e-commerce providers today. If this situation continues the way it is, with companies not paying much attention to customer service, the true promise of high-speed access will never be achieved. Broadband access needs to become a lifestyle, and this means that the complexity of DSL or other forms of broadband must be hidden from the consumer.

Establishing standards and resolving spectrum compatibility concerns are other items critical to a successful deployment of DSL. The lack of standards leads to market fragmentation, low volumes, and high cost. Moreover, use of standards-based DSL promotes the interoperability of technologies and devices, which is key to a successful rollout of DSL. The interoperability of devices also vastly increases customer choices, provides a greater selection of products, and reduces costs to the supplier and the consumer. Standards-based DSL also allows the market to provide more product and supplier alternatives to consumers than would have been possible without standards. Moreover, standards-based DSL helps focus the industry to resolve other challenges hindering its mass deployment.

Broadband access is one of the crucial pieces of the next-generation networks. Today companies are investing massive amounts of money to build and deploy next-generation backbone technologies. Hence, it is clear that the Internet and other packet-based networks will shortly have the necessary bandwidth and capacity to meet this explosive growth of data and video traffic. Companies are using various technologies and streaming techniques to move content closer and closer to the edge of the network, thereby avoiding the congestion that is currently slowing down the network. Edge devices are increasingly becoming smarter and smarter. What is lacking is access to the last or first mile, the residential area, and the small-to–mid-sized customer. Unless this existing bottleneck at the edge of the network is removed, next-generation networks will have massive core capacity with no way for consumers to access it. DSL is one of the access technologies that can be used to reduce, if not eliminate, this bottleneck.

As mentioned, there are nine types of DSL service:

1. ADSL
2. G.lite
3. Rate-adaptive DSL (RADSL)
4. ISDN DSL (IDSL)
5. Symmetrical DSL (SDSL)
6. G.shdsl
7. High-bit rate DSL (HDSL)
8. HDSL2
9. very-high–data rate DSL (VDSL)

The majority of all business deployments to date of any type of DSL have been made by data CLECs such as Covad, NorthPoint, and Rhythms (before bankruptcies). The incumbent local-exchange carriers (ILECs), notably SBC, have the majority of all residential deployments. Also, while the CLECs have focused on the less price-conscious business (e.g., the SOHO and teleworker markets), the ILECs have chosen to go after the truly residential consumer–a much tougher marketplace. This marketplace is extremely price conscious and will likely have a high degree of churn. More than 46 percent of the respondents to the broadband access survey thought that churn would be between 10 percent and 20 percent. Although currently churn is less than two percent, it will likely increase to 10 percent as deployment grows. The survey respondents predicted that within two years, 37 percent of all broadband users would shift from one type of access technology to another and drop their present provider in favor of another. The ILECs have focused on the consumer marketplace, for regulatory reasons and because of a fear of cannibalizing their T1 revenues.

Although the conventional voice circuit has only a 3.4-kilohertz (KHz) bandwidth, the physical wire connection's bandwidth is more than 1 megahertz (MHz). In the 1930s, when the U.S. government had decided to offer all Americans universal phone service, the most expensive parts of the network were the links connecting the COs. To keep telephony affordable, the industry decided to maximize the number of calls on each wire that went between the COs and the network. As such, they limited each transmission to 4 KHz. However, now that the United States has switched from analog switches to digital switches, the cost considerations that these guidelines were based upon are no longer necessary. DSL exploits this extra bandwidth to send data to the CO, where it connects with the company's fiber-optic network.

Much of the downstream bandwidth is unusable today because of the distances and length of the lines. DSL downstream rates depend on the length of the copper line, its wire gauge, the presence of bridge taps, cross-coupled interference, and the services provisioned on adjacent circuits. Signal attenuation in the copper and crosstalk are the biggest stumbling blocks for DSL. Technical solutions to the problems of larger dynamic ranges and longer distances are extremely complicated and expensive.