This is a guest post written by Robert Conger, director, product management at ADTRAN.
The need for speed continues unabated. To satisfy this need, service providers are deploying fibre deeper into the network and next-generation DSL equipment closer to the customer. This enables shorter loops and when combined with advanced DSL technologies, like VDSL2, allows for higher service rates. Unfortunately, these performance gains can be impacted by as much as 50 per cent by crosstalk. Without intervention, it is virtually impossible to deploy 100Mbps on a single access line.
Enter vectoring. Vectoring can virtually eliminate Far-End Crosstalk (FEXT) – the main source of performance limitations. With vectoring, Ultra Broadband service of 100Mbps can be deployed on a single copper pair as far out as 1,800 ft. (600 m) from the DSLAM. When two pairs are bonded, 100Mbps service can be extended to 3,400 ft. (1,100 m). This is a substantial improvement. However, for vectoring to be effective in the real world, a holistic approach known as system-level vectoring (SLV) is required.
A system-level approach is the only way to truly deploy vectoring in a scalable and cost-effective manner. This approach uses high-performance processing engines and very-high speed communication buses spread among VDSL2 line cards and chipsets. SLV also enables vectoring to work across cards in a system and between DSLAMs at the same location. A system-level approach of this nature ensures that rewiring is never required and that ports are never left stranded. Additional DSLAMs or cards enable ports to be added to vectoring groups with no subscriber ever seeing a sudden drop in performance – leading to an improved customer experience.
For a SLV approach to work, both DSLAMs and the VDSL2 chipsets they use must be designed with a system-level approach. Extensive digital signal processing is required to handle vectoring across a large number of ports.
First-generation vectoring-capable chipset solutions have taken different approaches toward SLV. Early adopters chose to move to either board level-vectoring or system-level vectoring but not both. Likewise, some equipment vendors chose to implement their designs as exclusively one or the other as well.
One way to ensure that separate DSLAMs do not transmit onto pairs in a common binder is to rewire customers so that each binder is handled by a single DSLAM. This approach, referred to as binder management, is generally problematic and expensive over the long term. It requires substantial ongoing work to continuously manage customers and leads to stranded ports. Additionally, inter-binder crosstalk has been shown to significantly reduce the overall vectoring performance improvement. SLV offers a distinct advantage. With SLV, DSLAMs have the ability to coordinate with each other, behaving as one system. This eliminates stranded ports and reduces costs.
Vectoring is a complex but effective new capability being integrated into next- generation DSLAMs. SLV enables service providers to rapidly deploy premium services without the cost and delays incurred when having to constantly rewire binders and the associated stranded ports.
To learn more about SLV, join ADTRAN CTO Dr. Kevin Schneider as he presents a talk entitled, “Vectoring De-Mystified” in the Access Evolution track on Day One of the Broadband World Forum 2012, taking place on the 16-18 October 2012 at the RAI Exhibition and Convention Centre, Amsterdam. Click here now to register your interest.