Deployment Considerations for FTTH in Multiple Dwelling

Deployment Considerations for FTTH in Multiple Dwelling
Deployment Considerations for FTTH
in Multiple Dwelling Units
White Paper
Optical
Fiber
WP1285
Issued: December 2007
ISO 9001 Registered
Ian Davis
Applications Engineer Manager - Europe
Introduction
Densely populated urban areas are being actively targeted for the next roll-out of fiber-to-the-home (FTTH)
deployments. Implicit in this enterprise is to capture as subscribers a large proportion of the 700 million
households occupying Multiple Dwelling Units (MDUs), i.e. flats and apartments, worldwide. Occupants of
MDUs are anticipated to represent the majority of urban residents eventually subscribing to FTTH.
Early investigations by operators into the feasibility of deploying optical cable into MDUs have unveiled new
constraints regarding installation techniques and deployments that must be taken into consideration. Although
standards have been created to address access network constraints, specifically ITU-T Recommendation
G.657, identifying two classes of fiber with improved macrobending performance, it is already apparent that
performance in excess of this standard will be necessary to enable successful roll-out of FTTH into MDUs.
This White Paper explains why performance beyond that defined in Recommendation G.657 is required to
achieve significant deployment cost savings and discusses the implications for fiber design.
Constrained Deployment in MDUs
Typically, FTTH within an MDU will involve an optical cable link being installed between the building entry
point and each subscriber within the MDU. The economics of maximizing ROI (return on investment) dictate
that installation should be achievable quickly, using relatively inexperienced technicians, and by minimizing
costs associated with “protecting” the fiber compared to deployments elsewhere in the network. This implies
the following conditions;
•
•
•
•
•
•
•
Flexible miniaturized cables capable of being maneuvred into corners and around obstacles to span the
link to the subscriber in the most convenient and visually unobtrusive manner.
Single operation to route the cable i.e. elimination of ducting to route the cable and provide protection
from external interference.
Rugged cable designs that can be treated just like copper cable, such as directly affixing the cable to the
walls of the building using fast, low technology equipment e.g. a staple gun obtainable from retail hardware
shops.
Some examples of the severe installation and routing found within buildings are presented in Figure 1. The
common feature of all these examples is that the optical fiber contained within the miniaturized cable becomes
confined in a bend radius radically smaller than is experienced by fiber in networks outside the building.
Figure 1. A montage of in-building installation and routing techniques that result in unusually tight bending
for optical fiber.
2
Improved Bending Fiber for Access Network Installations
ITU-T Recommendation G.657 attempted to address the concern that installation practices exemplified in
Figure 1 would lead to unacceptable levels of attenuation in conventional single-mode fibers complying to
ITU-T Recommendation G.652. Table 1 indicates how the macrobending requirements of G.657 are
tightened to address the more constrained environment of the access network. Improved macrobending has
been demonstrated using a number of optical fiber design modifications, illustrated in Figure 2, such as
increased core refractive index, reduced mode field diameter, depressed cladding and trench cladding.
Tightest Bend
Radius Specified
(mm)
Specified Bend
Loss 1550 nm
(dB)
Specified Bend
Loss 1625 nm
(dB)
G.652 Table D
30
(100 turns)
n/s
≤.1
G.657 Table A
10
(1 turn)
0.75
1.5
G.657 Table B
7.5
(1 turn)
0.5
1.0
Table 1. Tighter macrobending requirements demanded by Recommendation G.657 compared to
Recommendation G.652.
Core
Cladding
Cladding
Increase Core Refractive Index
Reduce Mode Field Diameter
Depressed Cladding
Trench Cladding Design
Figure 2. Alternative designs for improved macrobending performance to meet requirements of
Recommendation G.657.
3
Fiber Performance in Constrained MDU Deployments
Table 1 indicates that compliance to Recommendation G.657.B implies capability to deploy in bend radii as
low as 7.5 mm. However, as deployments in MDU started to roll-out it became evident that optical power
budgets were becoming immediately challenged, even though fiber complying with G.657 was deployed. On
investigation, it became apparent that the G.657 standard under-estimated the degree of bending that could
be imparted by the installation approaches described earlier when using the flexible and visually discreet
miniaturized cables that are regarded as ideal for within building deployment. Cables being bent through
90º into corners and around obstacles were found to be routinely imparting bend radii as low as 5 mm on the
constituent fiber. When the bending introduced to the optical fiber due to stapling an indoor cable directly
to the wall was considered, partial bends in the optical fiber of were also identified with the induced bend
radius also reaching 5 mm. The degree of bending induced by stapling was confirmed by x-ray analysis of a
cable affixed by this technique, see Figure 3. Bottom line, if miniaturized cable will experience 90º turns and/or
stapling, performance at 5 mm is absolutely required.
Figure 3. X-ray of bending of fiber in miniaturized cable due to direct affixing using stapling technique.
Even for fiber meeting the requirements of ITU Recommendation G.657, that has been designed using one
of the conventional techniques shown in Figure 2, macrobending losses down at 5 mm bend radii can be far
above the level that is tolerable for MDU installations. Figure 4 extrapolates the G.657 specification to predict
performance to 5 mm radius. At 5 mm radius, this extrapolation indicates bend loss per full turn of >10 dB
for G.657.A and >1 dB for G.657.B at 1550 nm. In comparison, specification for Corning’s ClearCurve™
optical fiber, a G.657.A, G.657.B and G.652.D product designed specifically to meet the challenges of MDU
deployment, is an order of magnitude improved, ≤1 dB/turn at 1550 nm. This represents a truly bend insensitive
macrobend capability that is ideal for operation in MDU applications.
4
Macrobend loss at 1550 nm (dB/turn)
10
1
G.65
7A
0.1
G.6
Clea
rCur
0.01
ve F
iber
57 B
TM
Capa
bility
0.001
0.0001
5
6
7
8
9
10
11
12
13
14
15
Radius (mm)
Figure 4. Extrapolation of ITU-T Recommendation G.657 performance to 5 mm and comparison with
Corning’s ClearCurve optical fiber.
Considering that the optical power budget within the MDU (after the intrinsic fiber attenuation, losses
for splicing and conectorisation and margin for maintenance and repair has been extracted) may be as little
as 0.5 dB for the type of multiple bending events illustrated in Figure 1, it is readily apparent that fibers
only just conforming to the G.657 standard will be unsuited to demands of the MDU FTTH deployment.
The necessity of deploying an optical fiber that not merely complies with the G.657 standard but exceeds
this performance and supports deployment to 5 mm bend radius was evidenced by a Corning study
comparing the impact of stapling an indoor miniaturized cable containing commercially available G.657.A,
G.657.B and G.652.D complaint fiber with a chemically doped trench cladding with the same cable design
containing ClearCurve optical fiber. Figure 5 demonstrates the increase in cumulative loss induced by
twelve incremental staples for the two fiber types.
Cumulative Impact of Staples on Loss Budget of FTTH Installation
0.5
Chemical Doped Trench Cladding
ClearCurve
Cumulative Loss (dB)
0.4
0.3
0.2
0.1
0.0
1
2
3
4
5
6
7
8
9
10
11
12
Stapling Event
Figure 5. Comparison of bend loss due to incremental stapling in miniaturized cables.
5
The chemically doped trench cladding fiber, as well as suffering a significantly greater average loss per
staple, has a much wider distribution of possible loss per staple, being susceptible to occasionally very severe
incremental loss events resulting in a “perilous” staple of >0.1 dB, a feature entirely absent from the
ClearCurve fiber which is designed and specified to provide excellent bending loss to 5 mm induced bend
radius. With typical links in MDUs from basement to subscriber featuring several tens of staples, not to
mention several 90º bends to navigate around corners and through walls, an installer is highly likely to
encounter enough perilous staples to challenge the power budget of a link with only 0.5 dB margin for
induced bend loss. Since optical signal testing would only follow the completion of the installation, and
would not be monitored staple by staple as in the described test, full rework of the imperilled cable would
be the probable outcome of applying the chemically doped trench cladding fiber to this application. Note
this study was performed using round staples that are sanctioned by some FTTH network operators. The
substitution by flat staples can lead to even more severe losses and the control of staple type in the field
may be difficult for the network operator to impose.
Consequently, the conclusion is drawn that even G.657.B macrobending performance is insufficient to meet
the stringent challenges of FTTH installation in MDUs. Enhanced optical performance to 5 mm bend
radius is a requirement for this application space.
Mechanical Reliability Considerations
Having established that 5 mm bend radius is a necessary requirement for fiber installed in FTTH installations
within MDUs, and that Corning’s ClearCurve optical fiber is optically capable of operation in this confined
deployment, it is also necessary to consider the mechanical reliability of fiber in this situation. Corning’s
White Paper WP1282, “The Mechanical Reliability of Corning Optical Fiber in Small Bend Scenarios”
describes in detail Corning’s use of a long established and accepted model for determining the failure
probability of lengths of fiber subjected to multiple stress events of either tension or bending. This model
is then applied to the particular situation of relatively short lengths of fiber constrained in very tight bends.
It is important to appreciate that only the length of fiber actually placed under bend is susceptible to stress
induced fatigue that may lead to risk of failure. For context, a full 360º turn of 5 mm radius consists of only
31 mm of fiber under bending stress. Since a failure over a 20 year lifetime requires the coincidence of a
partial tight turn (e.g. around a corner or induced by a staple) with a highly infrequent, low initial strength
flaw that fatigues to below the installed stress over the installation lifetime, it is perhaps unsurprising that
the failure predicted by the model per full 360º turn of 5 mm radius is just 3 parts per million. This failure
rate can straightforwardly be scaled up to the probability of failure per subscriber or per network by
considering the number of corners required to negotiate the typical path to the subscriber and the frequency
of stapling points necessary to affix the cable in place. Note this methodology assumes that fibers under
bend endure stress that is only generated by the bending of the fiber and no additional tensile stress is
present. Cable designs or installation practices that cause significant tensile load to be transferred to the
fiber as the cable negotiates corners can lead to elevated rate of mechanical failure and should be disallowed.
As the predicted failure rate is so low and the installed fiber in MDU is very accessible in case of failure
following installation, it is concluded that the approach of embracing miniaturized cables with fiber
performance exceeding the requirements of Recommendation G.657.B and allowing installation techniques
that constrain sections of the fiber to bends as low as 5 mm is fully justified.
6
Corning Incorporated
www.corning.com/opticalfiber
One Riverfront Plaza
Corning, NY 14831
U.S.A.
Phone: 607-248-2000
ClearCurve is a registered trademarks of Corning Incorporated,
Corning, N.Y.
Email: [email protected]
©2007, Corning Incorporated
7
Was this manual useful for you? yes no
Thank you for your participation!

* Your assessment is very important for improving the work of artificial intelligence, which forms the content of this project

Download PDF

advertisement