Innovation in Access - a strategic solution for upgrading the
copper
plant By adopting an access network systems perspective and forming a team with over 65 man-years working with the xDSL access
plant, Rim Semi has applied fresh thinking to the copper
broadband challenge. We offer a fundamental advance in DSP performance that directly enhances the carrier's ability
to upgrade his access network bandwidth using only a seamless
line-card / CPE upgrade.
Our goal is two-fold: keep the needs of the access network
engineer in our minds at all times, and make certain our DSP
fulfills the strategic business goals of the copper carrier.
The result is an innovation in xDSL performance that assures
the carrier can implement a quality of IPTV services that will
be cable-competitive.
Cable-Competitive Services means the
picture quality, capacity of channel content, and reliability
of triple-play services will surpass that of digital cable
plants and that copper carriers will not only be able to sign
up new subscribers, but can actually win subscribers back from
cable companies.
But
these business objectives will only be feasible when the needs
of the network planning engineer are fully realized.
According to a recent lab analysis by a world-renowned DSL
testing firm, the Cupria technology has achieved these technical
benchmarks, enabling carriers to deploy cable-competitive
services while reducing their access network costs by over
50% compared by VDSL2. Below is the systems perspective
we have engineered into our breakthrough DSP architecture
to enable delivery of our system goals.
The Importance of "Pure IP" Triple Play The importance of making a robust, reliable access
transport technology available cannot be overstated. The VP
in charge of network planning at one of the Big Three US
telcos told us directly that their business plan for broadband
over copper was in doubt because they were challenged in
making their 5Mbps DSL work. Their business plan for 3-tiered
DSL service was working fine for the bronze and silver
subscribers at 1.5 and 3 Megs respectively, but that service
interruptions at 5 Megs was intolerable. So if today’s DSL
transport at 5Mbps is not reliable, how could they
realistically expect to extend their access network to serve
30Mbps to feed an IPTV strategy?
IP-only environment One of our objectives in designing a broadband over copper
solution form the ground up was to provide an IP-only
environment free of the complexities of ATM.
We believe that
the clearly emerging trend
for an all-IP access network makes
the most
sense for both carriers and equipment
makers – and
provides the
quickest path to making
broadband over copper
as
reliable as POTS.
bit-error rates across the
access network determine the
ability to compete with cable services
However reliability alone is useless if payload
bandwidth capacity is not large enough to accommodate
the multimedia traffic that carriers need to become
cable-competitive. Our
target service capacity is 30Mbps of Ethernet payload. We
believe this is adequate to support the next 20 years of
domestic household requirements for voice, video, and data
throughout the Multimedia Home.
IPTV Work-Arounds Another way to describe our thinking on Cupria is to ask
the question, “If you were going to design a new
video-over-copper solution for the access network, would it
end up looking like today’s VDSL2?”
Repeatedly we have been told by carriers that no, VDSL2 is
merely a patched-up VDSL that still cannot deliver the
fundamental advances in bandwidth payload, copper reach, and
quality of service that a cable-competitive IPTV deployment
requires. In 2006, the dominant IPTV installations in
actual deployment are using ADSL2+ serving 18Mbps out to only
3Kft and similar numbers are being used for VDSL2 planning.
We believe a robust IPTV access network must be planned for
30Mbps serving 6Kft.
The
Network QOS Challenge for Triple-Play Originally the copper access network was engineered to
support voice intolerance for audible delay, and to
compromise bandwidth requirements with an acceptable level of
noise in the circuit. Emerging data requirements flipped this
priority in QOS for the copper loop – and ATM provided an
ideal method for assuring both data and voice QOS. However
today’s emergence of IPTV brings a third – and most demanding
– intolerance of both noise and delay. Ethernet data packets
will easily re-transmit and re-assemble to make up for
noise-induced packet loss. But streaming video cannot
tolerate IP delay, which results in picture break-up and/or
MPEG artifacts. Delivering a cable-competitive grade of
digital video service will mandate that the access loop be
engineered to deliver a Bit Error Rate of at least 10
-9 (which delivers an annoying picture artifact
every 6 minutes for a 3Mbps SDTV stream).
Cupria’s adoption of a Pure-IP transport approach help bridge
the gap in today’s network performance to make
cable-competitive IPTV both technically and financially
feasible for the copper carrier.
Our goal is two-fold: keep the needs of the access network
engineer in our minds at all times, and make certain our DSP
fulfills the strategic business goals of the copper carrier.
The result is an innovation in xDSL performance that assures
the carrier can implement a quality of IPTV services that will
be cable-competitive. 
According to a recent lab analysis by a world-renowned DSL
testing firm, the Cupria technology has achieved these technical
benchmarks, enabling carriers to deploy cable-competitive
services while reducing their access network costs by over
50% compared by VDSL2. Below is the systems perspective
we have engineered into our breakthrough DSP architecture
to enable delivery of our system goals. 
Originally the copper access network was engineered to
support voice intolerance for audible delay, and to
compromise bandwidth requirements with an acceptable level of
noise in the circuit. Emerging data requirements flipped this
priority in QOS for the copper loop – and ATM provided an
ideal method for assuring both data and voice QOS. However
today’s emergence of IPTV brings a third – and most demanding
– intolerance of both noise and delay. Ethernet data packets
will easily re-transmit and re-assemble to make up for
noise-induced packet loss. But streaming video cannot
tolerate IP delay, which results in picture break-up and/or
MPEG artifacts. Delivering a cable-competitive grade of
digital video service will mandate that the access loop be
engineered to deliver a Bit Error Rate of at least 10