VDSL (Very high speed Digital Subscriber Line) is the next generation of copper loop access. VDSL is strategically paired with Fiber-to-the-curb, and the pair delivers extremely high speeds into homes while utilizing the existing copper loop on premises. ADSL is called a "last mile" technology. VDSL is a last-hundred-feet technology. It is designed to work over very short distances at high speeds, so fiber has to deliver data up to the curb or a similarly close point, and VDSL takes it inside the house and to the phone, television, and computer.

VDSL is a hingepin of telecom's multimedia delivery strategy, because a single VDSL loop can deliver digital television, voice, and data into the home. It is similar to the digital fiber cable initiative (called hybrid fiber cable or HFC) that that cable companies are currently deploying, because it requires rerunning the network with fiber to the neighborhood.

VDSL is technically very similar to ADSL. In fact, it can be considered ADSL optimised for short distances. Its upload channel is asymmetric at 1.6-2.3Mbs (with future plans for 19.2Mbs and symmetric access), while its download is between 50-55Mbs (but can degrade down to 13Mbs on long runs). It runs in a separate frequency band from voice and ISDN, allowing POTS to run through the same line as data. It will use either CAP (Carrierless Amplitude Modulation), DMT (Discrete Multitone), or DWMT (Discrete Wavelet Multitone), most likely DMT though my vote is with DWMT.

While the fiber backbone will undoubtedly be ATM-based, the premises network will be difficult to design. The first method would be to create a CPE device that divides the television, voice, and data on to their respective networks. This simplifies upstream multiplexing and administration and in new installations it is ideal. However, in existing networks, it would be far easier to allow every device (computer, television, telephone, etc) to connect to the VDSL network on its own via a passive adapter. However this is most costly, more difficult to administer, and potentially degrading. It also implies designing the VDSL protocol for multiple access points, and vastly affects the resultant system (unlike issues like line coding which are political and meaningless).

Also, very little fiber has be deployed into neighborhoods, let alone to the curb, so VDSL may be a solution without a market until telecoms can get their act together and distribute fiber. So far, they have been watching cable companies putting heavy bets on HFC and generally losing at every step, so there is not a big initiative to deploy fiber. Further, most telecoms don't even have the backbone to support OC-1 speeds for every house and business in town anyway. It will likely be at least five years before VDSL makes its first big splash on the market.