Vectoring Regaining the Momentum of Copper

Why Vectoring?

To meet the growing demand for triple-play services—that is, combined data, voice, and video services—carriers have turned towards fast broadband connection. Their short-term goal is 20–30 Mbps per user, and their medium to long-term goal is 50–100 Mbps per user. FTTH is represented by GPON and can easily reach these speeds. However, it requires large-scale investment, and regulatory issues in last-mile deployment also need to be solved.

VDSL2 allows copper lines to provide FTTH-like bandwidth. This helps carriers significantly reduce their investment in broadband. Testing has shown that VDSL2 can reach speeds of up to 100 Mbps in both the upstream and downstream links. In field deployment, however, background noise, EMI, impulse noise, and crosstalk adversely affect VDSL2 bandwidth and transmission distance.

In short- and medium-haul VDSL2 scenarios (within 1 km), crosstalk in a copper binder is the most significant noise that affects VDSL2 bit rate. Crosstalk cannot be mitigated by new encoding or modulation techniques. As a result, finding other ways to mitigate or even cancel crosstalk is the key to widespread commercialization of VDSL2.

Vectoring mitigates crosstalk between twisted pairs in a big copper binder. With vectoring-enabled equipment, carriers can easily achieve bit rate that is comparable to crosstalk-free VDSL2. This has driven carriers to exploit their legacy copper resources and offer high bandwidth to users at low cost.

Since it was released in 2010 as the ITU-T G.993.5 standard, vectoring has matured and demonstrated its huge commercial potential. Vectoring is expected to be widely deployed in 2012.

What is Vectoring?

Crosstalk typically occurs as far-end crosstalk (FEXT) or near-end crosstalk (NEXT). NEXT couples between a receiving path and a transmitting path of DSL transceivers at the same end of two different subscriber loops within the same twisted-pair cable (Fig. 1). FEXT is detected by the receiver located at the end of the cable farthest from the transmitter that is the noise source.

VDSL2 is based on FDM, which can eliminate NEXT because upstream and downstream traffic is transmitted in different frequency bands, and a filter embedded in the transceiver can filter out all unnecessary frequency bands. However, FEXT cannot be mitigated by the filter because noise and information are transmitted over the same frequency band. Because VDSL2 loop length is typically shorter than 1.5 km, FEXT in VDSL2 is more severe than in other DSLs. FEXT is therefore the main type of crosstalk in VDSL2. FEXT decreases signal-to-noise ratio, which means line transmission rate decreases and error bit rate increases. This affects cable performance.

A dedicated vectoring control entity (VCE) is used to coordinate the processing of signals from all lines that exert crosstalk on each other in a cable binder. VCE cancels out FEXT through signal compensation so that bit rates are almost FEXT-free.

The VCE performs upstream or downstream vectoring signal processing or crosstalk cancellation processing on the central office side. Downstream and upstream signal processing requires different mechanisms.

In the downstream direction, all signals are precoded using a precompensation matrix algorithm before being transmitted from DSLAM. The final signals received by CPE are almost FEXT-free.

In the upstream direction, all signals are decoded using an offset matrix algorithm after being received from CPE. The final signals processed by the upper-level protocols in VTU-O are almost FEXT-free.

The coordinated signal processing mechanism, also called vector processing, treats multiple VDSL2 ports as a multiple input multiple output (MIMO) system to mitigate or even cancel input and output crosstalk between lines. Each MIMO system constitutes a vectoring group.

ZTE's Unique System-Level Vectoring Solution

VDSL2 equipment is suitable for deployment in a street cabinet/node (FTTC mode). When deployed 500 to 1000 meters from homes, a DSLAM covers about 200 homes in most cases.

Two mainstream vectoring solutions are available: board-level vectoring (BLV) and system-level vectoring (SLV). BLV processes VDSL2 ports on each board (usually 32 or 48 ports) as an individual MIMO system or vectoring group. This results in multiple vectoring groups in a DSLAM. By contrast, SLV treats the whole DSLAM as a vectoring group so that all ports are placed in one vectoring group.

Vectoring can only mitigate crosstalk between lines within one group but not crosstalk between different groups. Thus, SLV is the better choice for deployment.

ZTE's system-level vectoring solution has a unique design comprising a centralized vectoring process and separate vectoring and VDSL2 access. ZTE took the traditional architecture of control switching card and VDSL2 line cards as its basis and developed an innovative architecture comprising control switching card, vectoring processing cards, and VDSL2 line cards. With this architecture, carriers can choose either traditional VDSL2 access mode or vectoring and VDSL2 access mode to increase bandwidth and reach.

This architecture enables seamless evolution to a vectoring-enabled system. Previously, carriers tended to deploy traditional DSLAM using an architecture with control switching card and VDSL2 line cards. Now, carriers can add a vectoring processing card to seamlessly upgrade to a vectoring system without replacing any of the existing hardware. ZTE’s system-level vectoring solution has the following benefits:

• remarkable performance improvement. A lab test was conducted using a 32-pair, 0.4 mm cable over 400 m. When vectoring was enabled, line performance significantly improved. Any of the multiple pairs in a big copper binder can achieve transmission performance that is almost FEXT-free. Testing showed a maximum of 146.42% (31.29% on average) improvement in bandwidth downstream, and a maximum of 280.34% (56.57% on average) improvement in bandwidth upstream.
• reusing copper resources cuts costs. To attain 50−100 Mbps bandwidth, carriers tended to choose FTTH as their mainstream access mode. However, FTTH means reconstruction in the last mile, which requires huge investment. As vectoring matures, carriers have begun to realize that copper resources can be reused to achieve high bandwidth without costly reconstruction.
• seamless upgrade to protect investments. With ZTE’s unique vectoring architecture, carriers can first deploy a VDSL2 DSLAM with vectoring-enabled user line cards to provide VDSL2 access. As users demand more bandwidth, a carrier only needs to add a VCE card to make the whole system vectoring-ready without replacing any hardware.
• low power consumption. According to tests, vectoring significantly reduces the power consumption of VDSL2 ports. At the same maximum transmit power (Pmax = 14.5 dBm or Pmax = 2.5 dBm), a vectoring-enabled VDSL2 chipset performs much better than an ordinary VDSL2 chipset. In the case of a lower maximum power (Pmax = 2.5 dBm) for a vectoring VDSL2 chipset and higher maximum power (Pmax = 14.5 dBm) for an ordinary VDSL2 chipset, vectoring VDSL2 is still performs better than ordinary VDSL2. With vectoring enabled, the line driver can save up to 300 mW per port, while the additional vectoring unit (VCE) adds 100 mW per port. Overall, 200 mW of power is saved for each port.

ZTE unveiled its industry-first system-level vectoring prototype at BBWF Europe in 2010. The prototype was the center of attention. In 2011, ZTE was shortlisted for two Infovision awards: "Broadband Innovation of the Year" and "Broadband Access Network Technologies and Services-Fixed". Carriers worldwide are incorporating vectoring into their broadband development strategy because it is cost-effective and performs excellently.