By far, the most widely used broadband technology to access the Internet around the world today is DSL. Over 70% of broadband connections in Japan and Korea are DSL, while penetration in Europe is approximately 60%.
In the U.S., cable TV companies dominate broadband. But DSL isn't far behind, with about 46% of total connections. In fact, DSL growth is estimated at 15% to 20% annually, so it's expected to pass cable TV sometime this year as the dominant broadband technology in the U.S.
That says a lot, considering DSL uses the century-old #26 AWG twisted pair of the plain-old telephone system (POTS). Nevertheless, DSL is alive and well. In fact, it's expected not only to continue its growth in the U.S., but to become even better to meet the telcos' triple-play plans.
While growth figures for DSL are attractive, they don't tell the whole story. The Government Accountability Office report last year indicated that only 58% of U.S. households had any sort of Internet access, with 30% using dialup and 28% using broadband (cable, DSL, satellite). Another study from the Pew Internet & American Life Project showed that 42% of U.S. citizens have access to a broadband connection, the majority of whom reside in cities and the suburbs.
Rural areas are really underserved, with only an estimated 25% having any kind of high-speed Internet connection. (Yet AT&T and Verizon—the two largest DSL providers—claim to offer DSL service to 80% of their subscribers). This shortage won't be easily resolved, as the telcos and cable TV companies usually don't think it's profitable to provision so few customers in such remote areas.
Together, these factors leave a huge opening for some reasonably priced satellite technology or other wireless option. Satellite broadband connections are already available in some areas where satellite TV service is available. But the opportunity is a reality for DSL providers because the wiring is there. All they need to do is enable the service.
Flavors Of DSL
All DSL systems are based on orthogonal frequency-division multiplexing (OFDM), which divides the high-speed bit stream into many slower parallel streams and modulates each on adjacent carriers. The telcos refer to OFDM (also known as multicarrier modulation) as discrete multitone or DMT. Several versions are being implemented, and ADSL is the most widely used by far.
A Fresh Example
Texas Instruments' UR8 residential gateway architecture chip set exemplifies the latest trends and features based on DSL (see the figure). Designed for home gateway boxes, the chips cover all the latest DSL offerings, including the newest VDSL2 standard.
Also, these chips feature a multimedia gateway processor, a programmable DSL physical layer (PHY), a high-performance DSP-based voice subsystem, and a rich set of local-area network (LAN) interfaces. A well-defined application program interface (API) reduces time-to-market by allowing hardware and software reuse across all DSL platforms.
This chip set lets residential gateway designers handle Voice over Internet Protocol (VoIP), as well as multiple methods for distributing broadband services within the home. Some of the home networking technologies supported include Ethernet, Wi-Fi, and standards offered by the HomePlug Powerline Alliance, Multimedia over Coax Alliance (MoCA), and the Home Phoneline Networking Alliance (HomePNA).
No one knows which home networking technology will dominate in the future. Most homes with a network rely on Wi-Fi because of the conveniences of wireless. But will existing or even future wireless technologies have the bandwidth to support the triple play of high-speed data, VoIP, and Internet Protocol TV (IPTV) and video on demand (VOD)? That remains to be seen.
While IEEE 802.11g Wi-Fi systems seem to have the potential, current access points have limitations because walls, ceilings, and other obstructions shorten range and create dead zones. High attenuation and multipath cause speed to back off dramatically for a reliable connection, making video transport more iffy.
That's why so many vendors are joining with the MoCA and HomePNA crowd to support wired technologies that will virtually guarantee quality video transmission. Some new wireless systems such as the 802.11n upgrade and TZero's version of Ultra-Wideband (UWB) using multiple-input multiple-output (MIMO) may do the job. In any case, expect multiple home networking methods in the mix. Whatever works the best for the least money will win.
DSL is prospering. According to TI, there are 226 million subscribers worldwide as of 2006, and the latest projections claim an estimated 170 million more by 2010. New installations will be with ADSL2+ or VDSL2. Yet there are few ADSL2+/VDSL2 installations here in the U.S. Most telcos still sell and implement the original ADSL, which is good for 768 kbits/s to 1.5 Mbits/s, but that's it.
Faster systems have been implemented in some areas. But obviously, the cable runs are shorter. This means a smaller number of subscribers—those near the central office—as the only potential customers. Carriers have installed repeater boxes in neighborhoods to give some areas higher speeds. But these faster services aren't available to the bulk of the DSL customers.
With that limitation, how will the telcos offer HD, VoD, and IPTV? Some offer TV now through satellite TV deals, but what about IPTV over DSL? If the telcos can only reach a fraction of their DSL customers now with the higher-speed services, it doesn't look like a promising or profitable venture for them.
The main option for the telcos appears to be fiber to the home (FTTH). Already seeing this problem, Verizon has implemented its FiOS (fiber-optical service) to select parts of the country. Using a hybrid cable TV-like system over a passive optical network (PON), Verizon delivers super-fast data service that can easily handle IPTV. AT&T is testing its similar U-verse system in Texas using a PON to neighborhood terminals and then VDSL2 to the home over twisted pair into the home.
Fiber is the answer, but it is very expensive. When IPTV finally arrives, demand should increase, creating a bigger demand for FTTH and hybrid fiber-cable systems to deliver it. The telcos aren't likely to give up on delivering IPTV and leave that market to the cable guys. In the meantime, the DSL path is well defined if a profitable way to deliver its faster versions can be found.