"There's one big problem -- telecommunications bandwidth"

"Bandwidth bottleneck. No question, that's the biggest obstacle."

The incredibly rapid boom of the Internet use and the development of on-demand services as well as other multimedia applications have forced telecommunications industry to create solutions that would increase bandwidth, providing faster access to different networks and more efficient use of various applications. Furthermore, the competition in telecommunications has increased along with sharpened customer demands, as the number of players continues to grow due to more open markets worldwide. This paper explores one remote acces solution, that is xDSL technology from both technical and business point of view.

xDSL is a modem technology, enabling high-speed delivery of data, audio, and video in a digital form over the existing telephone infrastructure (copper twisted-pair phone lines called the local loop which connects the customer premises and the carrier’s central office). There are several types of DSL (Digital Subscriber Line) technology, namely DSL, HDSL, SDSL, ADSL, VDSL and RDSL. Various DSL technologies differ from each other in data rate, operating distance, ratio between downstream and upstream speeds, as well as concerning their applications.

The basic idea behind all xDSL technologies is that a special modem is attached to each end of the copper phone line, i.e. one modem at the customer premises and another one at the central office. The phone line is then exploited with the help of advanced modulating techniques. Furthermore, technological advances in digital signal processing, transformers, analog filters and A/D converters contribute to xDSL technologies.

This section discusses the plain DSL and the five different xDSL technologies and also highlights some of their applications. More attention is given to ADSL, which is currently the focus of interest and most sought-after among the xDSL technologies.

DSL modems create altogether three channels, two B channels and one D channel. The data rate of the D channel is 16 kbps and it is mainly used for signalling. Both B channels can carry up to 64 kbps duplex (downstream speed equals the upstream speed). The operating distance of DSL is 18,000 feet.

DSL is a modem used for narrowband ISDN. The target customers of ISDN include not only businesses but also residential subscribers. ISDN can be used for voice and data communications and its speed is sufficient to support also desktop videoconferencing. Several experts do not, however, see ISDN as a major end-to-end service to support multimedia applications. They rather predict that its main use will be to provide home access to the Internet. On the other hand, the increased use of real-time video and audio over the Internet necessitates even higher-speed access to the Internet.

HDSL is the most mature of the xDSL technologies. It has been fully standardized by ANSI (American National Standards Institute) and ETSI (European Technical Standards Institute).

HDSL requires two twisted pair lines to carry data at a rate of 1.544 Mbps both upstream (from the subscriber to the network) and downstream (from the network to the subscriber). Another alternative of HDSL technology uses three twisted-pair copper lines to carry 2.048 Mbps. These data rates are provided as far as 12,000 feet.

HDSL technology is mainly used by early adopters in a restricted area (such as student living in a campus area) as well as by private organizations, as it demands two or three twisted pair copper lines. Typical applications of HDSL include access to the following systems: PBX network, cellural antenna stations, digital loop carrier systems, inter-exchange POPs, Internet servers and private data networks.

SDSL has the ability of transferring data at the same speed as HDSL, but it requires only one twisted-pair copper wire. Additionally, POTS (Plain Old Telephone System) and digital data transfer can occur simultaneously on SDSL. However, the operating distance of SDSL is only 10,000 feet.

Typical SDSL applications are the same than supported by HDSL. However, SDSL has one important advantage over HDSL: it is also suitable for residental users, as they usually have only a single telephone line.

Applications that demand a higher data rate downstream than upstream can benefit from ADSL technology. With an asymmetric data stream, signal coupling is reduced and a higher data rate and a longer distance can be achieved.

According to the ADSL Forum, the following three channels (which can be used simultaneously) are created in the twisted pair line by interconnecting ADSL modems at each end of the local loop:

1. a high-speed downstream channel
1. a medium-speed duplex channel (includes both upstream and downstream)
1. a POTS channel, which is separated from the digital ADSL modem by filters

Server

CORE

According to ADSL Forum, the above mentioned three channels are created by dividing the telephone line with the help of either of the following methods. In both techniques, a 4 kHz region is reserved for POTS:

• Frequency Division Multiplexing (FDM), which assigns one band for upstream data and another band for downstream data. Both these bands are then divided into channels by time division multiplexing.
• Echo Cancellation, which first assigns the upstream band to overlap the downstream band and then separates the two by local echo cancellation.

According to Whitehouse Publications, Discrete Multi-Tone (DTM), a multi-carrier modulating technology, is generally used as the modulation standard for ADSL. DTM is very resistant to noise. An alternative modulation code for ADSL is the Carrier-less Amplitude/Phase modulation, CAP.

At the customer premises, the ADSL modem is attached to a PC, telephone or facsimile machine. Consequently, at the carrier’s central office modem, the voice data (POTS) is separated from the digital data and then forwarded to the public telephone network through the CO switch. Accordingly, the digital data in turn is forwarded to an Internet Service Provider, for example.

The downstream speed depends mainly on the distance and the wire gauge. A data rate of up to 9 Mbps can be reached with a wire less than 9,000 feet long and 24-26 AWG thick. If the distance is doubled, the data rate will fall to 1.544 Mbps. The upstream and downstream speeds on a duplex channel vary accordingly from 16 kbps to 640 kbps.

ADSL has a special feature, error correction capability, which enables ADSL to correct errors due to impulse noise. Moreover, ADSL is partly standardized by ANSI and ETSI.

Because of the asymmetric nature of ADSL, it is not the best solution for many important business applications requiring roughly equal amounts of bandwidth for incoming and outcoming data, as in videoconferencing and LAN interconnection. ADSL is most suitable for applications in which far more information is downloaded than uploaded. In future, home video-on-demand and other on-demand services, teleshopping, WWW browsing etc. might be used with ADSL technology.

VDSL transfers asymmetric data at rates much higher than ADSL, but over much shorter lines. According to ADSL Forum, the maximum downstream speed of 51.84 Mbps can be reached only if the copper wire is max. 1,000 feet long. With a distance of 4,500 feet, the data rate is 12.96 Mbps. The corresponding upstream rates are 2.3 Mbps and 1.6 Mbps respectively.

However, VDSL is still under development, and it is viewed that in the future VDSL might achieve the same data rates both downstream and upstream, but at a shorter distance. Moreover, there are no general standards for VDSL yet, but several standards organizations and forums are currently working on the standardization of VDSL.

VDSL is expected to be future residental access technology that supports similar applications than ADSL. VDSL technology could also be deployed in High Definition TV (HDTV).

RDSL is a rate adaptive xDSL technology, which means that it can work at different data rates depending on the length of the line and thus, it can overcome the distance limitations for xDSL to a certain extent.

RDSL allows speed adjustment according to the line quality. It enables carriers to serve users who reside a long way from the nearest central office. A remote user may not get the fastest speed because of the distance of the line, he will, however, get the best possible rate. Thus, RSDL can basically support all the same services supported by DSL, HDSL and ADSL depending on the line distance.

In the future, as xDSL technologies are deployed in high volume, there will be a need for a new way to funnel data from the modems to the backbone. Ethernet switches and routers designed for private corporate networks have been used for xDSL trials, but unfortunately they do not have enough capacity to satisfy the future needs. To overcome this, a device, called Digital Subscriber Line Access Multiplexer, DSLAM, is being developed. It should be able to aggregate data from xDSL modems and switch it to the backbone network.

In addition, according to Carol Wilson, it is likely that Asynchronous Transfer Mode (ATM) will be used with the xDSL technologies to switch data from the xDSL modems to the network trunks.

Given the rapid boom in the Internet use and users’ frustrations with slow access in the local loop, high-speed remote access is one of the promising applications for xDSL. According to International Data Corporation, the number of Internet users is expected to quintuple in few years, from approximately 40 million in 1995 to nearly 200 million by the year 1999. Accordingly, the growing sales of personal computers will contribute favourably to the successfulness of xDSL systems: it is estimated that by year 2001, another 240 million PCs will be sold. There is indeed market for xDSL: according to a Dataquest report, world sales of remote access systems exceeded USD $4 billion in 1996.

According to the ADSL Forum, the prospective market for ADSL has changed since 1989: although ADSL was originally designed for Video on Demand (VOD) applications, it is however, currently seen that the immediate market consists of users wanting proper service and willing to pay more money for high speed access to the Internet as well as for access to corporate networks from residences or remote branch offices. Therefore, the real profits from ADSL will be generated via personal computers.

Although ADSL is still in the market trial stage with telephone companies around the globe, many telcos intend to start offering ADSL services for Internet access and other packet-based communications during 1997. More than ten thousand ADSL modems and thousands of lines have already been installed and tested successfully in over 30 telephone companies especially in Europe and North America. HDSL, on the other hand, has been used by telephone carriers for years as a cost-effective way to provide T1 lines, enabling them to turn up services in hours rather than in weeks or months. More than 250,000 HDSL lines are already in place, and almost every T1 line installed during the past few years has relied on HDSL modems.

Although xDSL is one of the most promising alternatives for remote access, it will not immediately replace competing access technologies. At the moment, it is still quite difficult to estimate the size of the market for xDSL. Nevertheless, telcos are preparing themselves for high demand. For example, the Joint Procurement Consortium (a group of U.S. telcos: Ameritech, Bell South, Pacific Bell and Southwestern Bell) plans to deploy more than 2 million ADSL lines over the next 5 years. According to Risto Linturi, technology manager of the Helsinki Telephone Company HPY, expects that they will have about 10,000 xDSL users within their network area by the year 2000. These potential customers will be households requiring high-speed connections as well as enterprise telecommuters. GTE Telephone Operations, on the other hand, believe that if their ADSL trial will be successful, ADSL can expand their high-speed data portfolio and become an alternative to cable modems.

Several telcos classify the customers for xDSL technology in a similar way. US WEST, for example, is going to target ADSL/HDSL access towards high-speed Internet "power users", enterprise telecommuters, small business/branch office LAN interconnectivity and Internet Service Providers (ISPs). Thus, the following market segments will be briefly discussed in this section (see next page):

• Local Exchange Carriers (LEXs) and Inter-Exchange Carriers (IXCs)

The operating environment of today’s carrriers is changing drastically: increasing competition, surging customer demands for increased bandwidth at less expensive rates and congested voice systems are among the biggest challenges. DSL technology enbales carriers to respond to these challenges. Carriers can provide DSL services subscriber-by-subscriber on demand, and thus reduce their risk considerably. Moreover, DSL can be offered even in remote rural areas without substantial capital expenditures, because users can be added incrementally simply by adding modems.

Inter-exchange carriers (IXCs) have also begun to compete for customers in local markets. HDSL technology enables IXCS to offer, on the one hand, traditional service (T1 leased line) to their business subscribers and on the other hand, high-speed Internet connectivity to their own Internet Service Providers (ISPs).

xDSL technology can fill a market niche that is in high demand by small businesses, Internet "power users" and telecommuters requiring fast, affordable access to the Internet and enterprise networks. ADSL can, for example, connect widely dispersed business and residential users, and enable them to can apply ADSL technology for Video-on-Demand and other on-demand services, for teleshopping and high-speed Internet connections in the near future. Moreover, they could receive digital live television and virtual CD-ROM access with the help of VDSL systems.

In recent years, the number of small businesses and home offices has increased quite considerably and is likely to continue to grow. The spreading use of fax machines, data communications in commercial transactions and the Internet have made small, individual businesses significant users of data services. Carriers should not neglect the potential, although it most often resides in residential or ex-urban areas, creating wide geographic diversity. According to Lecoma International Ltd., small business/home office (SOHO) as a target group is the fastest growing segment worldwide: users are, however, price sensitive requiring low costs, and thus may terminate service quite quickly. Nonetheless, the use of data communications for credit transactions, EDI, database inquiries and especially the use of the Internet will increase steadily in volume, creating market for high-speed remote access and hence, for xDSL technology. The Internet and telecommuting applications for SOHOs certainly create market opportunity for ADSL in particular.

Another target group for xDSL are the business services with service locations geographically concentrated in the urban centre or in clusters around the area. They are high capacity customers, having multiple requirements for voice, data, video and LAN inter-connectivity. Business services contribute to the growing bandwidth requirements for Internet, EDI and videoconference and they share a common need to upgrade or even change services quickly and inexpensively. HDSL in particular is suitable for videoconferencing and other business applications that require regular, heavy symmetrical data and voice communications, whereas VDSL could be used for remote LAN access.

Nowadays companies need to deliver information and services far beyond the corporate campus, to scattered users in branch offices, various business partners, and a growing number of home-based telecommuters. Private data networks of government agencies or large multinational organizations can take advantage of HDSL to get more efficient, reliable, cost-effective method to interconnect each one of their buildings for communications. Corporate network, e.g. Intranet, can consist of headquarters, factory warehouse, offices and retailers. Finally, business customers for xDSL can also be retailers providing their customers with greater access to and use of the Internet, thus promoting their sales on the Internet.

The xDSL revolution allows traditional Internet Service Providers (ISPs) to differentiate themselves by providing subscribers with high-speed access to their networks and thus, ISPs can compete with giant telcos offering Internet services. ISPs can use gain a competitive egde by using HDSL to offer high-speed access to remote users.

For competitive access providers (CAPs), HDSL provides an attractive solution to target business customers that were previously inaccessible due to their distance from the fiber rings. CAPs can offer traditional business services, such as T1 leased lines, at a lower cost using HDSL technology while providing high-speed Internet connectivity to an ISP as well.

Telephony providers have an existing infrastructure to address billing, customer support and installation. Nevertheless, if Telcos are to deploy xDSL, billing systems will have to undergo some changes. One option could be flat-rate billing. Telcos could, on the other hand, bill xDSL services differently for different customers in order to establish quality of service: business customers are likely to be willing to pay more for higher level of service than a casual residential customer. Existing billing systems might not, however, be equipped to handle that kind of differentiation and need thus be reinforced, for example by using front-end processors to take the different parameters and translate them into something older biling systems can interpret. Telcos expect to charge about USD $40 to $100 per month for ADSL services. For instance, the !NTERPRISE unit of US WEST estimates that the initial pricing will be in the USD $60-100 range for unlimited service. Whatever the tariff for xDSL will be, telcos say that they intend to be competitive with Cable Modem pricing.

Different sources of information give very varied price estimations for xDSL. Some claim that the cost of the modems is in the USD $2,000 to $3,000 range, but telcos say they need to come down to about $500 per subscriber in order the ADSL to be profitable. Suppliers, however, argue that costs will come down as soon as telcos start buying ADSL gear in volume. In overall, xDSL modems cost anywhere from USD $1,500 to $3,000 per line. As xDSL technology becomes more widely applied in public networks due to more mature xDSL systems entering the market, the volume ordering by carriers begins to drive prices down. Lecoma International, Ltd. estimates that cost per connection for DSL and SDSL varies in the range of USD $500-1,500, whereas the pricing for HDSL can fall below USD $1,200 per system for large quantities. VDSL prices are not available at the moment, since only a few vendors can supply components. According to Lecoma, ADSL prices of equipment vary a lot due to the wide availability of the applications.

The following sections provide examples of companies operating with xDSL technology, namely operators, ISPs, techonology licensors and equipment suppliers.

At the moment, almost every major telephone companyes is researching the xDSL technologies in order to solve the problem of scarce bandwidth and thereby meet their customers’ demands for faster Internet connections and new services, e.g. various multimedia applications. xDSL technologies were supposed to be obtainable from most of the interested companies already in the mid 1997; it seems however, that mass deployment of xDSL will not occur until 1998-1999. Currently, ADSL is the most sought-after of the DSL technologies due to its suitability for Internet access. As a matter of fact, the ADSL forum with well over 200 companies worldwide as its members was established to coordinate the process of introducing new solutions based on ADSL technology. More than 40 telephone companies worldwide have conducted successful trials on xDSL technologies.

Most of the companies operating in the field of Digital Subscriber Lines are based in the United States. The next section reviews some of the major teleoperators in the U.S. and in Finland by giving examples of their achievements and views of the xDSL market.

There are two Finnish Internet service providers, i.e. Scifi and EUnet Finland, that presumably will have a central role as xDSL markets emerge in Finland.

There are only few companies licencing the technology needed in DSL products: Amati licenses ADSL and VDSL techologies; AT & T Paradyne ADSL, SDSL, and HDSL; and Aware Inc. licenses only ADSL. Amati and Aware have based their solutions on DMT techology whereas AT &T uses CAP technology.

Companies involved in DSL development can be divided into the following categories: semiconductor companies, modem suppliers, access network suppliers, network/service providers and consultants. Due to the availability of various technical applications and market’s early stage, a lot of emphasis has to be concentrated on customers easy access to new technology. Thus, many companies are cooperating and forming strategic alliances across categories. The next section presents some companies actually working with xDSL –products.

This chapter describes xDSL technology in terms of other access technologies and deals with the strategic gains xDSL can generate. A SWOT analysis is also provided.

xDSL technology has many advantages compared to other access technologies, such as analog modems, ISDN, cable modems using both cables coaxial and fiber, radio LAN, wireless cable (e.g. MMDS and LMDS) and satellite networks. This section briefly compares these alternative access technologies with xDSL.

The data rate of xDSL line is up to 51.84 Mbps (VDSL), whereas the data rate of the fastest existing analog modem is up to 56.6kbps. Compared to analog modem xDSL provides more possibilities for network content providers to create new services due to the high speed of xDSL in transferring data.

The data rate of ISDN is 128 kbps, i.e. much lower than the corresponding rate of xDSL. However, the need for fast data transfer is extremely large among residential power users of the Internet, home offices, large businesses, etc. The basic rate interface ISDN is not optimized for the increased use of real-time or "streaming" video and audio over the Internet, whereas xDSL is. Furthermore, in areas where demand for ISDN is low, xDSL will be the less expensive service to provision, as ISDN requires software upgrades and replacement of CO voice switching equipment, and telephone companies need a large number of users to cover those costs. Moreover, not all Internet Service Providers offer ISDN access, even though the local telephone company would provide ISDN.

As it could take up to 25 years to completely "cable" the U.S. with two-way hybrid fiber-coaxial (HFC) networks, fiber will is not likely to catch up with the millions of telephone lines lacing our planet for many years. xDSL provides an immediate broadband deployment option while these HFC networks are being built. Furthermore, cable modems use the cable network with 30 million subscribers at the moment, whereas xDSL uses the telephone network having already over 130 million subscribers. Furthermore, cable modems share a line with tens of other users and as more users join a line, the capacity available to any one user inevitable drops.

Certainly, fiber optic is the way of transferring data in the future, but it is still too expensive to install everywhere. On the contrary, in order to create a connection with xDSL modem, the whole neighborhood does not have to be cabled with fiber optics.

Disadvantage of radio LAN is the short radio range (indoors only up to 50 metres). On the other hand, xDSL’s range is up to 4,000m.

The cost of deploying satellite networks make prices relatively high compared with terrestrial options. Moreover, a satellite loop is not meant for replacing the existing local loops, but should be seen as complementary to mobile local loops.

This section analyzes the strengths, weaknesses, opportunities, and threats of xDSL technology.

-xDSL will be deployed over the existing network infrastructure. That makes it relatively simple, inexpensive and easy to deploy.

-xDSL does not require expensive and time-consuming switch upgrades.

-Advantage for service providers is that xDSL can be rolled out subscriber-by-subscriber on demand, reducing risk considerably.

-Telephony providers have an existing, comprehensive infrastructure to address customer support, billing and installation.

-Data privacy is a key advantage of xDSL technology.

-xDSL gives carriers a competitive time-to-market advantage in the lucrative remote access market. It will also enable them to offer advanced ATM and packet-based services to business customers over the existing copper access line infrastructure.

-xDSL provides both upstream and downstream transfer of data, needed for applications such as videoconferencing and on-demand services. These applications are not so well supported with, for example, the speed of ISDN as they are with the speed of xDSL.

-xDSL modems are still expensive for residential users

-The service area of xDSL technology is not very large ( up 4000 metres)

-xDSL gives an opportunity for service provides to create new services requiring high-speed data transfer in general and equal upstream/downstream transfer of data in particular.

-The lack of standardization: so far only HDSL fully conforms to standards set by ANSI and ETSI, whereas ADSL is only partly standardized by ANSI and ETSI.,

-Currently, there not so many content providers for services made available by xDSL and the question is "Who is going to pay for it?" Finally, the existing cable network and other services cable companies offer e.g. in the US might attract residential and business customers to invest in cable modems.

It is highly likely that xDSL will become widely adopted particularly among high-speed Internet "power users" as well business requiring fast access to various networks, provided naturally that the prices come down. This is certainly going to happen in Finland and elsewhere in Europe, as the replacement of copper lines with fiber will take too much time to provide a remedy or panacea for scarce bandwidth. Accordingly, the cable network especially in Finland does not support broadband cable modems at the moment. xDSL technology offers a viable interim solution to the bandwidth problem while fiber optic cables are being installed. The potential of xDSL technology and its advantages can be summarised with the words of Rick Gilbert, President at the ADC Kentrox.

ADC Kentrox, "The Need for Speed", http://www.kentrox.com/product/cellworx/nspeed/xDSLprimer.html

ADSL Forum "Cable Modems and ADSL", http://www.adsl.com/adsl/adsl_vs_cable.html

ADSL Forum, "General Introduction to Copper Access Technologies", http://www.adsl.com/adsl/general_tutorial.html

ADSL Forum, "VDSLTutorial", http://www.adsl.com/adsl/vdsl_tutorial.html

"Alternative local loop technologies" A Review, OECD Paris 1996, OECD/GD(96)181.

Amati, http://www.amati.com

"Business Applications Fuel HDSL Demand", High-speed Access for Business, http://www.kentrox.com/product/cellworx/hsaccess/hsaccess.5.html

Carol Wilson. "DSL Technology May Push ATM Into Service", Inter@ctive Week, February 10, 1997, http://www1.zdnet.com/intweek/print/970210/inwk0067.html

Cisco, http://www.cisco.com

"DSL Technology May Push ATM into Service", February 10, 1997, http://www1.zdnet.com/intweek/print/970210/inwk0067.html

L M Ericsson, http://ericsson.se/cobra

Fluckiger, F. (1995). Understanding Networked Multimedia. Applications and Technology. Prentice Hall, Hertfordshire.

Kim Maxwell: "ADSL Access Networks", http://www.adsl.com/adsl/adsl_access_ntwks.html

Microsoft, http://www.microsoft.com/isp/supercomm

Motorola, http://motorola.com

Nicholas C. Stanley, "Mining the Local Loop: New Services from Old Facilities", http://www.lecoma.com/tele96.html

Paradyne, http://www.paradyne.com

PC World, October 1996 in "High-speed Access for Business", http://www.kentrox.com/product/cellworx/hsaccess.intro.html

Rachael King. "Hurry Up & Wait" , October 1996, http://www.teledotcom.com/1096/features/tdc1096coverstory.html

Risto Linturi: "HPY tarkentaa multimediaverkon hinnoittelua", Letters to the Editor/Tietoviikko 21 February, p. 7 1997 (Information Technology newspaper)

Telechoice, "Status of North American xDSL Trials", March/April, 1996. http://www.telechoice.com/xDSLnewz/us.html

"The Carrier Competitive Advantage", High-speed Access for Business, http://www.kentrox.com/product/cellworx/hsaccess/hsaccess.4.html

"The Broadband Bonanza: Economical High-speed Remote Access over the Copper Infrastructure", High-speed Access for Business,