Telephone wire still good for 100 – 300 Mbps
Many people have written off copper telephone wiring as a dead end for next generation broadband, but Alcatel-Lucent has proven that the rumors of copper’s demise have been greatly exaggerated. The latest improvements in VDSL2 technology could see commercial deployments as early as 2011 at speeds of 300 Mbps within 400 meters from the Telco’s VRAD and 100 Mbps at 1000 meters. Most homes that fall within the footprint of AT&T and Qwest will be within the 1000 meter range so they could see service offerings around 100 Mbps or more.
The new technology uses a noise-cancelling signal processing technology called “vectoring” to bond four copper wires (standard telephone wire in homes) to achieve 300 Mbps at 400 meters. The technology could also be deployed in business applications with many more pairs of cables at speeds up to 2 gigabits. While 100 to 300 Mbps isn’t all that impressive compared to an all fiber access network, it works on existing telephone cable plant so it doesn’t cost several thousand dollars per broadband subscriber to run new fiber. With 65% of the world’s existing broadband subscribers running on copper wiring, this development is extremely useful.
Signal processing technology will continue to improve which would result in even higher performance from telephone wire. More importantly, 100 Mbps is well ahead of the content/application curve which is currently below 4 Mbps per user due to simple server-side economics. It is by no accident that this correlates to current average subscriber speeds of 3 to 6 Mbps in the United States. The correlation makes sense because customers have less incentive to subscribe to faster tiers if their is rarely any free legal content that absolutely requires it.
While VDSL2 can’t compete with the potential performance of an all fiber broadband network like Verizon FiOS (nothing comes close to fiber), it can compete with cable broadband. VDSL2 bandwidth is dedicated capacity for each home on the last mile. The bandwidth is usually shared with IP television service which would eat up an average of 6 Mbps per unique HD stream, but that still leaves a lot of capacity for broadband. This compares favorably to current Cable DOCSIS 3 offerings which typically share 80 to 320 Mbps between ~165 homes (assuming 500 homes passed with 33% uptake). However, Cable companies generally don’t need to worry about loop length issues or plant quality issues that limit a small percentage of their subscriber base to lower speed tiers because coax cabling is heavily shielded while telephone wire is unshielded.
Cable’s downside is that their capacity is shared between many homes while VDSL2 is dedicated, but the same downside is also their upside because they can dynamically shift capacity between each home and offer better peak and better average performance that VDSL2 based networks. So long as cable companies add capacity in response to growing traffic levels, they will likely be able to continue offering better average performance. Cable companies can allocate more 6-MHz channels to their broadband network (though they lose those channels for television or video on demand) and eventually bring their neighborhood capacity up to gigabit speeds. They can also break apart neighborhoods into separate DOCSIS 3 networks through a process known as “node splitting” and double the available capacity to each customer.
Broadband is competitive
All these improvements on VDSL2 and Cable cost money, but the good news is that competition between Cable and Telco is robust and active. Telephone companies had to upgrade because cable companies are dominating the broadband market and luring away phone service customers. In response, Comcast recently offered a free upgrade from 6 Mbps to 15 Mbps for their Cable broadband customers because of the competitive pressure from Verizon FiOS and the copper based VDSL2 networks.