Академический Документы
Профессиональный Документы
Культура Документы
net/publication/224216266
CITATIONS READS
57 533
5 authors, including:
Some of the authors of this publication are also working on these related projects:
Design and experimental evaluation of Raman optical amplifier amplification, funded by Natural Science and Engineering Research Council of Canada, PARI
program View project
Long Reach/Range Passive Optical Network for Next Generation Hybrid Wireless and Optical Access and Fiber-To-The-X/Home (FTTX/FTTH) Networks: National
Program of Science & Technology, KACST, KSA View project
All content following this page was uploaded by H. Fathallah on 23 March 2014.
purposes, there is a growing need for the moni- hence increasing PON reliability and reducing
An automatic toring of the PON fiber plant. PON monitoring operational expenses.
technology automatically identifies and localizes The most important issue in PON monitoring
monitoring tech- faults of the in-service PON optical infra- technology is cost, including capital expenditure
nique allows the structure. In doing so, it provides the NMS with (CAPEX, i.e., the initial cost of the monitoring
network operator to enhanced optical infrastructure visibility in real technology per customer) and operational expen-
time, thus speeding up the detection and local- diture (OPEX, i.e., the cost of system mainte-
detect faults without ization of faults. Monitoring avoids the opera- nance). The reason is that the PON market is
resorting to in-field tional expenditures (OPEX) and large service highly cost-sensitive, especially for the compo-
restoration times of offline troubleshooting, thus nents not shared between customers, such as dis-
technicians or relying enabling wider service differentiation and tributed monitoring nodes. Therefore, an
on customer stronger QoS guarantees. In addition, it paves expensive technology, even though it may pro-
equipment or feed- the way to potentially enhanced physical layer vide in-service full visibility of the optical infra-
protection mechanisms. structure to the network operator, may not be
back. This feature is Accordingly, PON monitoring has been interesting for PON applications. Consequently,
highly desirable as receiving increasing attention, and a variety of the monitoring technology requires simple
proposals have emerged [3, 5]. To accommodate design, fabrication, and implementation proce-
the deployment of the demand for monitoring technology, the ITU- dures to minimize the cost.
in-field personnel is T L.66 (2007) Recommendation standardizes the Capacity, in terms of the number of PON
usually equated with criteria for in-service maintenance of PONs. It branches or distribution fibers that can be simul-
reserves the U-band (1625–1675 nm) for mainte- taneously monitored, is the second desired fea-
increased PON nance and lists several methods to implement ture. Candidate monitoring technologies should
downtime and PON in-service maintenance functions such as be able to support at least the maximum split-
OTDR testing, loss testing, and power monitor- ratio of current PON standards (e.g., 1:128 for
OPEX. ing (i.e., monitoring a proportion of the signal ITU-T G.984 GPON). Accommodating larger
power). split-ratios increases the number of supported
Note that PONs need to be tested during customers, thus amortizing the expenses of the
installation to ensure that all fiber links and service provider and generating higher benefits.
components are properly installed and working. The monitoring technology should thus be scal-
Therefore, link characterization and diagnosis able in order to enable seamless and continuous
during network installation is also of great upgrades of the PON infrastructure (i.e., PON
importance and can easily be performed using capacity, reach, and customer base) at low costs.
one of the aforementioned testing methods. The simplicity of the monitoring architecture and
However, there is a growing need to monitor components directly affects the cost, and is
fiber link failures and degradations without dis- hence an important requirement. In addition, as
turbing ongoing services. In this article we focus for any maintenance and protection mechanism,
on the monitoring of in-service live PONs (i.e., reliability is primordial. Furthermore, to operate
after installation), where a service interruption in-service, the desired monitoring technology
due to monitoring is not permissible. should act transparently to the data band signals
In this article we review and compare the such as the L and C bands. Therefore, strict iso-
major optical-layer PON monitoring proposals, lation between the data band and monitoring
and address advantages and challenges of the signals is required.
monitoring techniques for deployment of high-
capacity PONs. In the next section we enumer- AUTOMATIC AND CENTRALIZED MONITORING
ate the desired features and major requirements An automatic monitoring technique allows the
of in-service PON monitoring techniques. We network operator to detect faults without resort-
then briefly review the basic principles of OTDR ing to in-field technicians or relying on customer
for point-to-point monitoring, and outline the equipment or feedback. This feature is highly
challenges and limitations of standard OTDR in desirable as the deployment of in-field personnel
PON (point-to-multipoint) applications. Non- is usually equated with increased PON downtime
OTDR-based techniques are then addressed. and OPEX. Besides, it allows the operator to
We particularly focus on two recently proposed enhance customer satisfaction by potentially
techniques: Brillouin frequency shift assignment reacting to faults before service disruption (e.g.,
and optical-coding (OC)-based reflection moni- through automatic protection switching [APS]).
toring. We also address in detail the advantages A fully automatic monitoring system is usually
and disadvantages of each of the mentioned centralized, allowing the NMS, from its location
techniques in PONs. Finally, we discuss promis- in the central office (CO), to remotely acquire
ing solution paths before concluding in the final complete live network information without
section. requiring the collaboration of customers or their
ONUs, as does traditional OTDR in a point-to-
point link.
REQUIRED FEATURES OF Both centralized and distributed approaches
PON MONITORING TECHNOLOGIES have been proposed for monitoring the fiber link
quality of a PON [3–5]. In distributed (decen-
GENERAL REQUIREMENTS tralized) monitoring strategies, active modules
By definition, an effective monitoring technology are placed inside the ONUs to measure perfor-
should be able to both detect a fault and provide mance and report to the NMS. These modules
the NMS with useful information for root cause periodically evaluate the uplink for a specific
analysis. Useful monitoring information enables fiber branch and may be implemented electroni-
technicians to perform fast network repair, cally at the ONU.
fiber branch end, the integrity of a specific large network sizes. In fact, due to the huge
branch can easily be investigated. The OTDR is splitter loss at the RN, it is difficult to extract
exploited for a full characterization of the corre- useful information from the OTDR trace beyond
sponding fiber branch. The shift in the power the RN. In addition, as the network size increas-
level of the reference reflection for a desired es, the selection of an optimal delay line
branch provides useful information for the becomes more challenging, and the complexity
OTDR trace analysis. Checking the stability of of the OTDR trace increases.
the strong reflection (located well above the
noise level) is faster and easier than analyzing Multi-Wavelength Approach — One other
the OTDR trace, and these reflectors are often simple approach would be employing a multi-
used as a first fault indicator in most OTDR- wavelength source and an arrayed waveguide
based techniques. grating (AWG) at the RN. This reduces the
In RR-OTDR, the choice of the fiber lengths PON monitoring problem to point-to-point link
requires an important trade-off between OTDR characterization, as illustrated in Fig. 3. In this
sensitivity and resolution. The required fiber case the tunable multiwavelength OTDR source
length is proportionally related to the transmit- should be very stable for reliable monitoring.
ted OTDR pulse width as well as the relative Isolation between the monitoring and data sig-
distances between the customers. While for very nals will be more strict than single-wavelength
short pulses small fiber lengths are required, the OTDR. In addition to its high cost, this tech-
OTDR sensitivity is very poor, limiting allowable nique also has limited capacity due to practical
splitter size at the RN. For longer pulses, sensi- limitations and very poor spectrum efficiency [9].
tivity improves. However, significantly long delay Its scalability is hence very low. Nevertheless,
lines are required, leading to lower OTDR accu- this approach provides a centralized monitoring
racy and larger dead zones (i.e., the area of an system that enables the NMS to both detect and
OTDR trace where events are not distinguish- localize faults.
able).
The NMS requires updated information on Electronic Solutions — Note that the function-
the customer distribution in the network; other- ality of an OTDR device can be implemented
wise, customer relocations cause false alarms. within the ONU at the customer side [10]. This
The RR-OTDR scheme does not scale well with approach, known as embedded OTDR, leverages
the electronics at the ONU for a cost-efficient
solution, such that embedded OTDR within the
ONUs becomes an integral part of the monitor-
ing network. In this scheme the monitoring seg-
ment transmits an OTDR trace from the ONU
CO upon request of the NMS at the CO when the
corresponding ONU is idle over the upstream
OLT RN channel. Therefore, this solution relies on in-
Feeder fiber band upstream signaling. As mentioned earlier,
RM
ONU
this solution is inadequate when a fiber cut hap-
OTDR pens, as all data and control channels linking the
NMS to the ONU are disrupted.
reasons, this technique is very unlikely to be A particularly attractive feature of this solu-
adopted commercially. tion is the self-configuring nature of the net-
work. Encoders are installed at the drop fiber
OPTICAL-CODING-BASED PON MONITORING ends without concern for the fiber length from
OC exploits signal-coding techniques (inspired the remote node, unlike RR-OTDR. Signal pro-
by optical code-division multiplexing) for con- cessing at the receiver differentiates returns even
trol- and management-layer signaling operations. for remarkably similar fiber lengths (within
In OC-based PON monitoring, passive out-of- meters). Customer relocations can be accommo-
band encoders (Encn) are placed at the extremi- dated without a re-allocation strategy based on
ty of each PON distribution fiber to identify and previous installations.
monitor it, as shown in Fig. 5a [12]. The data One of the challenges in evaluating any moni-
and monitoring signals occupy separate wave- toring solution is predicting system capacity as it
length bands (Λ d and Λ m, respectively) consis- varies with the specific topology of the PON,
tent with emerging standards. An optical source whether legacy or greenfield. Simulations of specif-
at the CO transmits the out-of-band pulses ic topologies can be performed; however, they do
downstream; an optical or electronic receiver at not probe the generality of the solution. Statistical
the CO processes the aggregate upstream reflect- examinations of topologies can provide outage
ed signal. probabilities for the monitoring system in general.
The encoders both reflect and imprint a Several research topics remain to bring this
unique code (i.e., specific to the PON branch) technology to the marketplace. Compact low-
on the source pulses. Waveband separators split cost periodic encoders are essential. While previ-
the data and monitoring wavebands at the ONU ously proposed fiber delay lines are simple, mass
and the OLT. Alternatively, a combination of in- production is problematic. An integrated solu-
line encoders and monitoring band-stop filters tion for the encoder would reduce both cost and
may be used at the branch termination points bulk. Signal processing challenges also remain to
prior to the ONUs, as is the case for RR-OTDR. increase the coverage capability of the decoding
The use of simple fiber Bragg gratings (FBGs) algorithm. A reduced complexity maximum like-
directly inscribed at the termination of drop lihood receiver has been proposed in [14], but is
fibers may be regarded as a particular case of nonetheless suboptimal and may leave room for
OC-based PON monitoring. Although the sim- performance improvement.
plest approach, the use of FBGs as wavelength The use of time- or wavelength-domain
reflectors shares the low scalability and band- reflectors to identify PON branches, as in the
width efficiency drawbacks of the multiwave- reference-reflector and wavelength-based OTDR
length technique described earlier. Nevertheless, approaches, may be treated as particular cases of
the use of in-fiber FBGs is particularly attractive OC-based PON monitoring [15]. Compared to
since it reduces monitoring power losses by wavelength-domain reflection monitoring, code-
removing the requirement for waveband separa- domain reflection monitoring trades its more
tors at the termination of drop fibers. complex reflectors for higher scalability and
While several encoders and receivers have bandwidth efficiency. Compared to time-domain
been proposed, the most cost-effective and high- reflection monitoring, it avoids the use of delay
performance solution that has emerged is a com- lines to differentiate branch fiber lengths and
bination of periodic codes [13] and an electronic offers potentially higher scalability, particularly
receiver [14], illustrated in Figs. 5b and 5c. Peri- in the context of future long-reach PON (LR-
odic codes were developed exclusively for this PON) applications. Moreover, the extension of
application, and have low loss, low-complexity OC-based monitoring to LR-PONs may be facil-
hardware, and good performance. Previously itated through the use of in-line reflectors [15].
proposed encoders based on optical orthogonal However, this places additional strain on the
codes exhibited much lower performance for this more stringent power budgeting constraints of
application [13]. code-domain monitoring.
OC-based monitoring does not require forklift ing OTDR- and non-OTDR-based proposals for
upgrades of the PON distribution infrastructure, PON monitoring, and address the practical chal- Hybrid techniques
as does BOTDR. Consequently, the design of lenges facing their potential deployment. Rather
simplified and more cost-effective encoder and than being exclusive, OTDR and alternative should be investigat-
system architectures is a promising research technologies such as reflection-based monitoring ed as promising
direction [13]. Although OC-based monitoring are complementary. Therefore, hybrid tech- solutions for deliver-
does not offer complete fault localization (only niques should be investigated as promising solu-
fault identification on the branch), it is potential- tions for delivering the maintenance and ing the maintenance
ly more scalable than BOTDR. Therefore, it is protection functionalities required by current and protection
well suited as a component within a hybrid solu- and next-generation PONs. OC-based methods
tion, discussed in the next section. are particularly attractive to implement reflec- functionalities
tion monitoring in the context of increasing required by current
PON sizes. and next-generation
SOLUTION PATHS TO SCALABLE
REFERENCES PONs. OC-based
COST-EFFECTIVE PON MONITORING
[1] M. P. McGarry, M. Reisslein, and M. Maier, “Ethernet Pas- methods are
Our review of proposed optical-layer PON moni- sive Optical Network Architectures and Dynamic Band-
toring technologies reveals two distinct and com- width Allocation Algorithms,” IEEE Commun. Surveys &
particularly attractive
plementary monitoring principles: standard Tutorials, vol. 10, no. 3, 3rd qtr. 2008, pp. 46–60. to implement
[2] F. Effenberger et al., “Next-Generation PON-Part III: Sys-
reflectometry and the use of dedicated reflectors tem Specifications for XP-PON,” IEEE Commun. Mag., reflection monitoring
at the termination points of distribution fibers. vol. 47, no. 11, Nov. 2009, pp. 58–64.
While reflectors are capable of speedy identifica- [3] K. Yuksel et al., “Optical Layer Monitoring in Passive in the context of
tion of faulty distribution fibers, they lack any Optical Networks (PONs): A Review,” Proc. ICTON ‘08,
2008, pp. 92–98.
increasing PON sizes.
accurate fault localization capability. Conversely, [4] D. Anderson, L. Johnson, and F. G. Bell, Troubleshoot-
whereas standard reflectometry methods are ing Optical Fiber Networks: Understanding and Using
inefficient for distribution fibers, they are capa- Optical Time-Domain Reflectometers, Academic Press,
ble of yielding highly accurate fault localization 2004.
[5] F. Caviglia and V. C. Biase, “Optical Maintenance in
in point-to-point settings. In addition, adapting PONs,” Proc. ECOC, Madrid, Spain, 1998, pp. 621–25.
standard reflectometry techniques for PON [6] EXFO, “Application Notes 110 and 201”;
applications is neither economical nor practical, http://exfo.com.
particularly due to their lack of scalability to [7] I. Sankawa et al., “Fault Location Technique for In-Ser-
vice Branched Optical Fiber Networks,” IEEE Photonics
larger PON sizes. In contrast, fault detection via Tech. Letters, vol. 2, no. 10, Oct. 1990, pp. 766–69.
the monitoring of reflected signals is potentially [8] L. Wuilmart et al., “A PC Based Method for the Localization
simple, cost-effective, and reliable. and Quantization of Faults in Passive Tree-Structured Opti-
Therefore, we expect that comprehensive cal Networks using the OTDR Technique,” Proc. IEEE LEOS
Annual Meeting, vol. 2, Nov. 1996, pp. 121–23.
monitoring methods will integrate both the [9] M. Thollabandi et al., “Tunable OTDR (TOTDR) Based on
aforementioned monitoring principles. To do so, Direct Modulation of Self-Injection Locked RSOA for
it is necessary to break the monitoring proce- Line Monitoring of WDM-PON,” Proc. ECOC, Brussels,
dure into two separate steps, whereby fault Belgium, Sept. 2008.
[10] W. Chen et al., “Embedded OTDR Monitoring of the
detection and the identification of faulty distri- Fiber Plant Behind the PON Power Splitter,” Proc. IEEE
bution fiber is carried out in real time through LEOS Symp., Benelux Chapter, Eindhoven, Netherlands,
reflection monitoring, and precise fault localiza- 2006, pp. 13–16.
tion is implemented subsequently through [11] N. Honda et al., “In-Service Line Monitoring System in
PONs Using 1650 nm Brillouin OTDR and Fibers with
OTDR. Individually Assigned BFSs,” IEEE/OSA J. Lightwave
In currently deployed PONs, the implementa- Tech., vol. 27, no. 20, Oct. 2009, pp. 4575–82.
tion of reflection-based monitoring will enable [12] H. Fathallah and L. A. Rusch, “Code-Division Multiplexing
faster troubleshooting, as they will allow the for In-Service Out-of-Band Monitoring of Live FTTH-PONs,”
OSA J. Optical Net., vol. 6, no. 7, July 2007, pp. 819–29.
NMS to exclude customer equipment malfunc- [13] H. Fathallah, M. M. Rad, and L. A. Rusch, “PON Moni-
tions as the cause for a loss of signal while indi- toring: Periodic Encoders with Low Capital and Opera-
cating the faulty distribution fiber. Technicians tional Cost,” IEEE Photonics Tech. Letters, vol. 20, no.
equipped with high-resolution OTDR can thus 24, Dec. 2008, pp. 2039–41.
[14] M. M. Rad et al., “Experimental Validation of Periodic
be dispatched immediately for exact fault local- Codes for PON Monitoring,” IEEE GLOBECOM ’09, Opti-
ization and root cause analysis. Hence, fiber cal Net. Sys. Symp., Honolulu, HI, Dec. 2009, paper no.
plant degradation may be detected long before ONS-04.6.
transmission errors occur or services fail. In [15] K. Fouli, L. R. Chen, and M. Maier, “Optical Reflection
Monitoring for Next-Generation Long-Reach Passive
future PON deployments where protection is Optical Networks,” Proc. IEEE Photonics Society Annual
expected to play an increasing role, reflection- Meeting, Belek-Antalya, Turkey, Oct. 2009.
based monitoring may be integrated with the
protection schemes as triggers for implemented BIOGRAPHIES
APS mechanisms, leading to reduced downtimes MOHAMMAD M. RAD (mehdi.mansouri@gmail.com) received
and higher quality of service. both his B.S.E.E. and M.S.C. from Sharif University of Tech-
nology in 2003 and 2005, respectively. In September 2006
he joined the Department of Electrical and Computer Engi-
CONCLUSIONS neering, Center for Optics, Photonics, and Lasers (COPL),
Université Laval as a Ph.D. candidate. His research interests
Cost effectiveness and scalability are among the include fiber-optic communications, long haul data trans-
major requirements for in-service monitoring of mission, multiple access networks, network monitoring,
PON fiber infrastructures. OTDR requires costly and sensor networks.
architectural enhancements to deliver fast auto- K ERIM F OULI is a Ph.D. student at Institut National de la
matic fault localization in PON tree topologies. Recherche Scientifique (INRS), Montréal, Canada. He
In this work we review some of the most promis- received his B.Sc. degree in electrical engineering at Bilkent
University, Ankara, Turkey, in 1998 and his M.Sc. degree in (2001–2002) at Intel Corporation creating and managing a
optical communications at Université Laval, Quebec City, group researching new wireless technologies. She is cur-
Canada, in 2003. He was a research engineer with rently a professor in the Department of Electrical and Com-
AccessPhotonic Networks (Quebec City) from 2001 to puter Engineering at Université Laval performing research
2005. His research interests are in the area of optical on wireless and optical communications. Her research
access and metropolitan network architectures with a interests include wavelength-division multiple access using
focus on enabling technologies. He is the recipient of a incoherent sources for metropolitan area networks; analysis
two-year doctoral NSERC Alexander Graham Bell Canada of optical systems using coherent detection; semiconductor
Graduate Scholarship for his work on the architectures and and erbium-doped optical amplifiers and their dynamics;
performance of optical coding in access and metropolitan and in wireless communications, optical pulse shaping for
networks. high-bit rate ultrawide-band systems (UWB), as well as per-
formance analysis of reduced-complexity receivers for UWB.
HABIB FATHALLAH [S‘96, M‘01] received a B.S.E.E degree (with She has served as associate editor for IEEE Communications
Honors) from the National Engineering School of Tunis in Letters and on several IEEE technical program committees.
1994, and M.A. and Ph.D. degrees in electrical engineering She has published over 70 journal articles in international
from Université Laval in 1997 and 2001, respectively. He ini- journals (90 percent IEEE/IEE) with wide readership, and
tiated the use of Bragg grating technology for all-optical/all- contributed to over 100 conferences. Her journal articles
fiber coding/decoding in optical CDMA systems. He was the have been cited over 750 times per the Science Citation
founder of AccessPhotonic Networks (2001–2006). He is Index (SCI).
currently with the Electrical Engineering Department, Col-
lege of Engineering and Prince Sultan Advanced Technology MARTIN MAIER (maier@ieee.org) is an associate professor at
Research Institute of King Saud University (Riyadh, KSA), the INRS. He was educated at the Technical University of
and adjunct professor with the Electrical and Computer Berlin, Germany, and received M.Sc. and Ph.D. degrees
Engineering Department of Université Laval. His research (both with distinctions) in 1998 and 2003, respectively. In
interests include optical communications systems and tech- the summer of 2003 he was a postdoc fellow at the Mas-
nologies, metro and access networks, optical CDMA, PONs sachusetts Institute of Technology (MIT), Cambridge. He
and long reach PONs, FTTH, network monitoring, and was a visiting professor at Stanford University, October
hybrid fiber wireless (FiWi) systems. 2006 through March 2007. He is a co-recipient of the 2009
IEEE Communications Society Best Tutorial Paper Award.
L E S L I E A. R U S C H [S‘91, M‘94, ‘SM‘00, F‘10] received a His research activities aim at rethinking the role of optical
B.S.E.E. degree (with honors) from the California Institute networks and exploring novel applications of optical net-
of Technology, Pasadena, in 1980, and M.A. and Ph.D. working concepts and technologies across multidisciplinary
degrees in electrical engineering from Princeton University, domains, with a particular focus on communications, ener-
New Jersey, in 1992 and 1994, respectively. She has experi- gy, and transport for emerging smart grid applications and
ence in defense, industrial, and academic communications bimodal fiber-wireless (FiWi) networks for broadband
research. She was a communications project engineer for access. He is the author of the book Optical Switching Net-
the Department of Defense from 1980–1990. While on works (Cambridge University Press, 2008), which was trans-
leave from Université Laval she spent two years lated into Japanese in 2009.