Amazing….Bell Labs Pushes DSL Speeds to 300 Mbps


DSL Phantom Mode Explained. Click to view the graphic.

In this era of fiber to the home, it’s easy to dismiss copper-based DSL, the broadband connectivity technology commonly sold by phone companies worldwide. Looks like it’s too soon to completely write off this technology, however. Alcatel-Lucent, a company whose lineage is as old as the phone itself, says its research arm, Bell Labs, has been able to achieve downstream speeds of about 300 Mbps (over a distance of 400 meters) or 100 Mbps over a distance of one kilometer.

This is possible using a technology called DSL Phantom Mode. According to the company:

At its core, DSL Phantom Mode involves the creation of a virtual or “phantom” channel that supplements the two physical wires that are the standard configuration for copper transmission lines. Bell Labs’ innovation and the source of DSL Phantom Mode’s dramatic increase in transmission capacity lies in its application of analogue phantom mode technology in combination with industry-standard techniques: vectoring that eliminates interference or “crosstalk” between copper wires, and bonding that makes it possible to take individual lines and aggregate them.

The idea behind DSL Phantom Mode is that incumbent phone companies such as AT&T (s T), which are heavily invested into the aging copper infrastructure, can keep using those pipes for a lot longer. However, phone companies such as Qwest (s Q) will need to install new gear in the central office and in consumer homes.

Alcatel-Lucent’s efforts aren’t the only attempts to extend copper’s life and make DSL go faster and faster. Stanford University professor John Cioffi is working on a gigabit DSL solution and has started a company to give DSL a lift.


Stephen Cooke

Alcatel-Lucent is not the only one looking at putting multiple 100’s of Mb/s over the existing copper plant… Have a look at When you can provide this kind of bandwidth over the existing – non-re-conditioned – copper plant, why would you rip up the streets and people’s back yards to install fiber at greater than 10x the cost?

Alex Leverington

@kirk I guess that I should say that DSL in a POTS system is analog because there isn’t full control over the spectrum whereas with Fiber, carrying an analog transmission is an exception to the rule. DSLAMs are usually installed alongside POTS whereas a fiber plant (for FTTH) is digital from end to end. With DSL, having to deal with attenuation + POTS requires more overhead and thus, more hardware and power.

With respect to ILECs, a fully digital FTTH environment is much more likely than a fully digital DSL environment.

Splicing is irrelevant; copper or fiber, the labor’s going to be the same (unions!).


yo Alex Leverington: fiber is not natively digital. fiber optics use [insert Dr. Evil finger quotes] “lasers” to send photons down glass fiber. The modulation of the frequency of the “lasers” is interpreted as digital pulses by optical transceivers at either end of the light pipe.

Copper uses bouncing electrons.

The problem with copper is that over the centuries many splices have appeared since back-hoes and weekend gardeners chop telephone cables up like spaghetti. Since you can make ugly splices in copper and still make a phone call, they still make ugly splices today.


It’s nice to hear of these technological advances, but, how does one incentivize a moribund bankrupt supplier to step up to the plate and enable any improvement to their existing (ageing) infrastructure?


not impressed really. Too expensive and only a stopgap solution. Far better to use fibre instead of the copper, and this bonding solution will only work for people close to the exchange who will have to pay for two phone lines to get the service. Protecting the copper cabal really does have to stop…

Alex Leverington

DSL technology has been capable of 100mbps for years. No matter how fast DSL gets, the phone lines must be reconditioned to be “clean” enough for the given speed. Second, because DSL is based on an electrical current and not light pulses, DSL speed improvements typically have a lower yield than fiber. This is because fiber is natively digital instead of analog which means fewer DSP components and a lower signal-to-noise ratio. DSL technologies like Phantom are possible due to Moore’s law and when you apply the same level of computing power to fiber you get far more speed.

In a nuthshell – DSL isn’t cost effective. It’s a play for keeping ILEC boundaries. Extending the life of DSL extends the life of an ILEC as a Monopoly and let’s them return the last “bit” of revenue out of those copper lines. Faster DSL is a political benefit for incumbents, not a competitive edge or cost saving measure.

Good to see you following broadband. Considering recent posts regarding AT&T, I hope this comment is found to be relevant.


Of course it is amazing but what about the speed of the rest of the infrastructure. If the servers are not able to use the available bandwidth….

Comments are closed.