Original article: Alcatel-Lucent’s VDSL2 vectoring comes of age
As the telecommunications industry converges in Paris for the Broadband World Forum, it would be natural to look for evidence of further advances in FTTH rollouts and technologies. But even as FTTH technologies gain market share, shipments of DSL equipment continue to be strong.
In fact, after a decline in 2007–09, shipments are hitting new highs. We forecast that there will still be almost 400 million DSL subscribers five years from now, including many on hybrid copper-fiber networks, and additional in-building VDSL2 in FTTB networks.
Recently, there were a number of VDSL2-vectoring-related announcements by equipment and software vendors. But Alcatel-Lucent’s commercial launch of its VDSL2 vectoring solution is a first. Alcatel-Lucent claims the solution has the capability to support 100Mbps at 400 meters.
Vectoring can fulfill promise of true VDSL2 speeds
VDSL2 shipments in particular, while still a small proportion of shipments, have been growing steadily and in 2Q11 crossed the five million port mark. The result is due to a number of ongoing FTTN and FTTC/Cabinet type of deployments where part of the loop between the central office and the home is replaced by fiber and then ADSL is upgraded to VDSL2 to take advantage of the higher potential throughput on shorter loops.
But cross-talk between one copper pair and another, as well as other sources of noise, reduce that theoretical throughput considerably.
Standardized by the ITU in April 2010, vectoring is a type of noise-cancellation technology that helps regain some of the VDSL2 line’s lost throughput. DSL line bonding, vectoring, and “Phantom DSL” are the three ways that can potentially boost copper network speeds, of which bonding is already being implemented. AT&T is bonding its VDSL2 lines in the US for extended reach. We don’t expect Phantom DSL to be commercially available before the end of the 2012–13 timeframe. These options will help service providers to utilize existing copper access infrastructure, with partial fiber deployment, but before taking fiber all the way to the home.
Almost every major broadband hardware equipment vendor has some type of vectoring product in the works. Back in 2009, Ericsson was the first to publicize and announce a live lab demonstration of vectoring over six bonded copper pairs and achieving 500Mbps, or approximately 83Mbps per pair at 500 meters. Subsequently, Nokia Siemens Networks, Huawei, ZTE, and Alcatel-Lucent all made similar types of announcements. ECI Telecom has been working on its own solution for years and is in discussions with potential customers.
But Alcatel-Lucent is the first to announce commercial availability of some products for its ANSI DSL portfolio, with the rest, including products for its ETSI portfolio, to be available in December and over 1Q12. Alcatel-Lucent has said it has been field trialing vectoring with several service providers [since] 2010; these include Belgacom, Telekom Austria, Swisscom, Orange, among others. According to Alcatel-Lucent, many varieties of DSL CPE are upgradeable via software, negating the need for large-scale CPE replacements.
Node-level vectoring not always useful
In October 2010, ZTE announced an SLV or system-level-vectoring prototype, the ZXDSL 9836 compact DSLAM supporting up to 192 VDSL2 lines. In mid-August 2011, Huawei announced its own NLV or node-level-vectoring prototype that would achieve 100Mbps over 300 meters.
Alcatel-Lucent has covered its bases by launching BLV (board-level vectoring or within a line card), SLV (system-level vectoring or across multiple line cards in a system), and what it calls multi-system vectoring, which would support vectoring for up to 384 lines across eight line cards in 2012 with support for four line cards in 2011. But Huawei claims that node-level vectoring, equivalent to multi-system vectoring, would be most effective. In such a scenario, vectoring is applied to all cables connected to a specific site, not just one system alone.
[However] node-level or even system-level vectoring may not be useful in Western Europe where we expect the most demand for vectoring. The European Union regulator’s open access requirements still remain a hurdle. A cable binder shared between multiple service providers, and a mix of vectored and unvectored lines, negates the benefits of vectoring.
It is possible that such requirements may eventually not be implemented due to practical reasons or that unbundling at the bit stream access level is recommended as it was in Australia. Otherwise, service providers may need to content themselves with active management of non vectored lines in order to maintain the gains of vectoring.