Ericsson is on a bit of a roll these days and is north of $9 after dropping below $7 in January. Clearly, an important element of Ericsson’s resurgence is its solid positioning in 3G and 4G wireless - the announcement of an important role as an equipment provider in Verizon’s 4G LTE network is definitely a sign of this. However, the news out of Ericsson this morning is that they are demonstrating a new fixed line technology called “vectorized VDSL” (henceforth referred to as VVDSL), which will provide a data transfer rate of 500Mbps over a 500m distance, or roughly 100x faster than average home Internet connections in North America. Why is this an important announcement and what market is Ericsson trying to address with this new product? Let’s dig into VVDSL…
What is VVDSL? In short, telcos are delivering broadband Internet connections through a number of avenues: Clearwire in the U.S. is delivering relatively low-speed connections over wireless using WiMAX (and pre-WiMAX standards), cable operators are delivering it through coaxial TV cables using a technology called DOCSIS, and Verizon FiOS is delivering through fiber lines using passive optical networking (PON). For legacy telcos however, those that don’t go the route of laying expensive fiber optic cabling end-to-end are using DSL technologies, which re-use the old copper cabling already going to houses, apartment complexes, and businesses, though fiber is generally brought to the neighbourhood and the signal is then split off to DSL to reach the last mile. And as companies like AT&T look to compete in a broadband war as cable, fiber, and wireless companies ramp up the speeds to offer more HD video and higher Internet speeds, next-generation DSL will be critical in keeping up with the pack.
While DSL has the advantage of leveraging a huge installed base of existing cabling, the problem is that the quality of said cabling isn’t great (much worse than cable’s infrastructure), and DSL technology needs to account for crosstalk between the lines especially as the rates go up. Ericsson’s new VVDSL technology looks to bring 5x the speed of the fastest current DSL technology called VDSL2, which was standardized in the 2006 timeframe. How will this be done? The technology behind VVDSL is channel bonding, which looks to take 6 different existing copper lines, and multiplex the data stream to get 5x the throughput. Sounds easy? Well, not really. As seen in this article from back in early 2007, bonding even two channels together may only increase speeds by 25-30% if the two copper channels are too close too each other. As noted earlier, cross-talk between the lines is the problem here, and Ericsson’s VVDSL technology adds important cross-talk cancellation capabilities to the mix.
The other big problem with DSL is the distance from the central office, and once you get more than a few kms away, the speed drops off markedly. In fact, the current top end DSL technology doesn’t perform any better than the previous gen ADSL2+ beyond about a km and a half. Channel bonding seems like a natural evolution of the technology, but so far, the upgrade to two-channel bonding has been slow for AT&T, though it’s desperately needed because for some consumers, their links are maxed out as soon as you run 2 HD channels into the home along with an Internet connection (probably looking at around 20Mbps here). However, the speed equation changes dramatically with VDSL2 if distances can be brought under 0.5km, and speeds of 50Mbps per user (or 100Mbps for pair-bonded links) should be sufficient for the next 5 years or so. At the end of the day, VVDSL isn’t all that exciting in the last mile for homes, in my view, except for AT&T being able to advertise “Up to 250Mbps to the home!” When looking at broadband to buildings and businesses, that’s a completely different story altogether.
The real application for VVDSL is in circumstances where installing fiber cabling would be prohibitively expensive and there’s sufficient demand. Pricing in the fixed broadband access arena might be getting more challenging going forward, especially with cablecos completing their DOCSIS 3.0 upgrades (which can offer more than 300Mbps downstream and 100Mbps upstream) and Verizon upgrading their fiber offering to GPON = 2.5Gbps. Ultimately, these companies are building out their backbones with fiber cabling, but VVDSL could help in providing to the flexibility to pick and choose when fiber goes into existing areas.
Perhaps the biggest driver of a VVDSL technology isn’t better end user access but as backhaul for wireless. With more consumers moving away from land lines to wireless plans, it may not be that long before we see the same transition take place in the business world. A visit to Ericsson in Stockholm a few years ago drove this point home, as employees there had already started to use their cell phones as a primary business extension. As the move to wireless (and to smartphones) takes place in both the consumer and business arenas, the importance of backhaul (providing the connection back to the telecommunications backbone) obviously increases. Providing cheap, high speed, and low latency connections back to the Telco provider is going to be the name of the game, and this year we’ve already seen a huge move towards femtocells – mini wireless basestations that go into consumer homes, which then route wireless traffic over a consumer’s broadband connection. For the service provider, this is economical because wireless spectrum is extremely valuable and for the consumer, they can save some money and/or get better performance over their iPhones.
While it’s a question mark how fast femtocells will propagate, major service providers are definitely getting their offerings ready (the value seems to be a bit more slanted towards the service provider because carriers want to pass off the cost of the femtocell on to the consumers). For DSL-based service providers, VVDSL may be able to provide really cheap broadband access and cheap wireless backhaul, using the existing copper infrastructure. In a sense, the volume of smartphone data going over the femtocell network may not strictly need this big a pipe, but these data connections need lots of headroom because of the real-time nature of services like VoIP – i.e. as soon as connections get congested, very bad things happen to time-sensitive connections. And with video teleconferencing likely to be an important service over the coming 5 years over video terminals and smartphones, moving this data over underutilized copper lines might be the cost effective way to go for a giant like AT&T.
The bottom line for the DSL world is that with VVDSL, 500Mbps is a decent amount of throughput for companies and apartment complexes especially at a time when service providers are looking to backhaul wireless connections over fixed lines. Longer-term, AT&T might have been better off going right to fiber when Verizon decided to take the plunge, but now firmly committed to DSL, Ma Bell has to pick and choose when and where it will deploy fiber. While VVDSL won’t be standardized until later this year, this VVDSL technology is a sign that Ericsson is looking at how it can help service providers navigate the waters as fixed and wireless telecommunications standards overlap and ultimately converge towards the day when all forms of telecommunications is delivered over one high-capacity fiber optic cable going into the home.
The end game for Ericsson is winning in 4G wireless, and I see that innovating on the wireless backhaul side of the equation will help Ericsson sell the entire solution to service providers… sooner rather than later.
