VIAVI Solutions

White Paper

CCAP and Remote

PHY in the Headend
Solutions for today’s dynamic cable networks

Consumers are changing the way they use video and data in the home.
Multicast, or broadcast programming, is no longer meeting consumers’ appetite
for content while demand for unicast services, such as video on demand (VoD), has
exploded. Part of the change is the result of consumers having more devices on which to
consume data and video.

Televisions are still the most dominant screen for family viewing, but individuals are increasingly using laptops,
mobile phones, tablets, and other Internet protocol (IP)-connected devices as well. With more screens in each
home, cable providers must find ways to keep up with bandwidth demand. In addition, the IP network is rapidly
becoming more dynamic in nature due to this increased unicast traffic.

According to a recent white paper by Cisco, global consumer Internet video traffic will be 69% of all consumer
Internet traffic in 2017, up from 57% in 2012. In addition, VoD traffic will nearly triple by 2017: the equivalent of 6
billion DVDs per month.

For cable providers, this new dynamic is causing traffic congestion in the headend of the network—the area
that links the plant with the consumer. Most cable networks still treat data and video as two separate entities,
yet as network traffic continues to grow in unpredictable ways, cable providers must look for new solutions to
handle traffic.

This paper explains the converged cable access platform (CCAP) approach as one such solution. It presents the
benefits of CCAP along with two possible implementation scenarios. The paper also discusses testing, as the current
testing environment will dramatically change. Finally, we will look at remote PHY as a way to further decrease
congestion in the headend.

Topics include:

y What is CCAP?

y Implementation

y Testing for CCAP

y Remote PHY

y Conclusions
What is CCAP? number of ports and individual QAMs that will have to
be tested/measured as part of install commissioning and
In a typical headend configuration, an edge quadrature
on-going maintenance will bring its own challenges.
amplitude modulation (edge QAM) unit handles
broadcast and narrowcast traffic. A separate cable
modem termination system (CMTS) handles data Implementation
and unicast traffic. However, with today’s dynamic Cable providers look at two different models for
network environment, the two separate systems implementing CCAP across their networks—integrated
must fight each other for bandwidth as demand CCAP and distributed CCAP. Which method they choose
grows. It is a challenge to manage all network traffic depends on how much traffic is currently running
effectively with two isolated systems in place. This through each headend and how much of that traffic is
causes congestion and bottlenecks that create quality broadcast and narrowcast versus unicast and data.
of service (QoS) issues for residential and commercial
customers alike. The Integrated CCAP method removes both the edge
QAM and CMTS units at the same time. It replaces them
CCAP eliminates this problem by combining edge QAM with one CCAP chassis that handles packet processing
and CMTS into one platform. It intelligently looks and data management and has line cards to handle both
at the amount and type of traffic flowing over the downstream and upstream traffic. This type of CCAP
network and makes adjustments in real-time to ensure also has modules to handle traditional HFC network
that traffic flows smoothly. In essence, CCAP puts all traffic and commercial EPON deployments.
downstream carriers, including DOCSIS and video QAMs,
in a single radio frequency (RF) port with the flexibility This all-in-one approach is cost effective in the long run
of changing between carriers via software. This creates and makes the headend as efficient as possible. However,
a transport-agnostic network architecture that deals it has significant upfront costs that may deter some
effectively with today’s traffic needs and will be providers.
compatible with future networks as they evolve.
The distributed CCAP model keeps the CMTS in place, but
CCAP helps consumers gain the bandwidth they need, replaces the edge QAM with a newer, high-density edge
when they need it, while also creating cost savings for QAM unit. This gives the headend more bandwidth and
providers. For example, CCAP exploits more bandwidth the ability to handle more narrowcast traffic, but does not
because it uses digital instead of legacy analog channels. address the dynamic nature of unicast traffic.
This adds five to six digital channels for every one
This approach offers providers more flexibility on
analog channel it replaces and uses more of the available
implementation and can save money upfront, but
spectrum within the cable.
costs more in the long run. A CCAP core chassis, that
CCAP also saves providers money since one unit can also manages the high-density edge QAM unit, will
now do what two units, the edge QAM and CMTS, eventually need to replace the CMTS.
used to do. This frees up as much as 50% of the rack
Which method a provider chooses ultimately comes
space in the headend, saving 50% or more in power
down to the situation they face at each individual
consumption.
headend. For urban areas with lots of traffic of all types,
CCAP edge devices also save space by having a higher the integrated approach is best. For more remote areas
port density. Each port can deliver a full lineup of QAM that do not have as much unicast and data traffic, the
carriers, with a single CCAP port serving a smaller distributed approach makes more sense.
number of nodes: ‘narrower’ casting. However, the sheer

2 CCAP and Remote PHY in the Headend

Testing for CCAP However, as more bandwidth is used, current methods
for testing noise and ingress detection become
CCAP solves many of the bandwidth and congestion
impossible to complete. This is because testing typically
problems cable providers face in the headend, but
takes advantage of unused bandwidth—testing from
testing and solving service disruptions become much
outside the channel to detect if noise is leaking out of
more challenging. In a pre-CCAP environment, providers
or into a specific channel. In a CCAP environment, all
manually created a channel lineup plan to determine
channels may be in use. New, in-band testing methods
which network programs to run over the physical
such as “ingress under the carrier” use an active channel
channels in the network. For example, if a consumer
and demodulate its traffic. These methods obtain
experienced pixilation while watching HBO, a video
data based on the modulation and by matching and
technician called the office to find out on which physical
determining what data doesn’t belong. This does not
channel HBO was broadcast. From there, he or she
interfere with the delivery of the traffic itself.
tested that channel to find the issue.

A similar problem
A CCAP environment creates the channel lineup New, in-band testing
is interference from
automatically, changing dynamically to best utilize methods such as “ingress
nearby wireless (over-
overall bandwidth. This becomes an issue for testing. under the carrier” use
the-air) networks such
The channel lineup may be completely different from an active channel and
as LTE/4G services
the time a customer experiences a problem to the time demodulate its traffic.
that operate in the
the technician conducts the test. The channel may even Providers can quickly
750 MHz range. As
change while a test is being conducted; this makes locate which channel is
the FCC and other
it even more difficult for a technician to know which having a problem, see if
regulators continue
physical channel has a problem. noise is leaking into or out
to sell spectrum,
of the channel, and correct
Another issue is that most cable providers have separate wireless carriers and
the problem without ever
technicians for video, data, and the physical layer— cable providers must
needing to move traffic.
the cable itself. In a CCAP environment, all three are use more frequencies
intertwined. Technicians must become experts at all which overlap with the
three and have completely integrated testers to handle LTE/4G range. Currently, cable providers move traffic
this new environment. from one channel that is experiencing interference to
a channel farther away to solve the problem. However,
Next-generation test instruments (such as the VIAVI as the network uses more channels, this solution no
Solutions VSE-1100) are now available that combine longer works—and the problem is only getting worse as
tests for data, video, and the physical layer into one LTE/4G rollouts increase.
fluid operation. They also auto detect the current
channel lineup so a technician can quickly test the In the age of CCAP, testing not only needs to be faster.
correct channel as soon as they are notified of a It also needs to help pinpoint where a problem is
problem. This is possible because these new test geographically located down to certain segments within
instruments are portable and robust enough to go out the network. For example, the increase in the number of
to the plant, see where the traffic currently is, and see QAMs being broad/narrowcast means that traditional,
what specific carrier or QAM is handling that traffic. slower test tools will take too long to complete a
This eliminates a lot of diagnostic layers and work for comprehensive quality check at the headend/hub for
the technician, saving valuable time as they try to solve every QAM. Test tools must speed up significantly in
a problem. order to decrease total test time for commissioning,
maintenance, and troubleshooting. In addition, the

3 CCAP and Remote PHY in the Headend

latest HFC monitoring solutions let the headend Remote PHY will come in the form of new node
monitor each node in the field and determine if a node hardware, replacing existing nodes in the field. It does
is unhealthy. New analyzers can then segment the line the same job as existing nodes, but adds dynamic traffic
between the headend and home to determine which allocation to the downstream and upstream channels.
segment is having the problem. Backhaul from node to the headend also changes
from RF optics to gigabit Ethernet optics in order to
One of the most frustrating interference issues is deliver higher capacity while using standard IP/Ethernet
impulse noise. This transient noise only disrupts a technology.
channel for a fraction of a second but causes significant
QoS issues for the consumer. Impulse noise itself is How soon can remote PHY be deployed? The main
hard to detect, but when complicated by the CCAP factor gating deployment seems to be finalizing a
dynamically changing the channel lineup, finding the technology roadmap that creates optimal, future-
root cause of a problem becomes even more complex. proofed, scalable solutions. A logical place for providers
The service experiencing the problem will be changing to start is in new, greenfield neighborhoods that don’t
and potentially different from one CCAP port to another. currently have nodes in place.

To counteract this issue, new testing units need to not

Conclusion
only auto-detect the current channel lineup. They also
need to make the impulse noise visible and show it The question is not if CCAP will be implemented, but
graphically so technicians can decide on the best course when. Demand for bandwidth is growing at such a
of action to resolve the problem. rapid pace, cable providers have no choice but to deploy
some version of the solution. CCAP looks like a wise
investment as it has the ability to be compatible with
Remote PHY
future networks as they evolve.
With all the positive changes CCAP is bringing to the
headend, one problem still remains as consumers CCAP also offers different methods of implementation
continue to use more bandwidth—up and downstream to meet the needs of each cable provider. It lets a
capacity. Remote PHY brings downstream QAM provider select a CCAP model that meets their needs
generation, upstream demodulation, and more on a headend-by-headend basis. This provides the
processing closer to the home to increase capacity. flexibility to implement CCAP in the most cost-
Put another way, remote PHY moves part of the effective way to bring down cost-per-bit and increase
headend/hub out into the field (at the node), closer to bandwidth.
people’s homes.

Nodes used to handle traffic to and from as many as

3000 homes, but with the increasing bandwidth and
service demands from consumers, that number is now
closer to 250 homes per node.

Contact Us +1 844 GO VIAVI © 2020 VIAVI Solutions, Inc.

(+1 844 468 4284) Product specifications and descriptions in this
document are subject to change without notice.
To reach the VIAVI office nearest you, ccap-wp-cab-nse-ae
visit viavisolutions.com/contacts. 30175977 900 0914

viavisolutions.com