Ethernet technologies have contributed to the lowering the cost of FTTH network
implementation. Application of the Ethernet standard has expanded from the enterprise
LAN to the access network environment. However, the application of Ethernet
technologies to the access network systems is still at an intermediate stage
of development (Fujimoto, 2006). In order to support
the services applications a number of network technologies are proposed, each
with advantages and drawbacks. Before going to transport solutions in the access
market it must noted that the actual transporting is done by a bandwidth, which
refer to not only the quantity of data it can carry, but the speed at which
the signal travel; bandwidth is measured in megabits or gigabits per second.
In a point to multi point (P2MP) PON, the downstream 1490 nm wavelength and
upstream 1310 nm wavelength are used to transmit data and voice. The downstream
wavelength 1550 nm can be used for analog video overlay. Multiple Optical Network
Termination (ONT) are connected through one or more optical splitters. In a
point to point (P2P) system, the voice and data are transmitted on the same
wavelength downstream and upstream because it uses a fiber pair; one fiber downstream
and another one upstream. Because the downstream power is divided among the
ONTs by the splitter, FTTH P2MP PONs have considerable optical loss. In order
to ensure that each branch of the P2MP PON will operate correctly and meet all
specification, the network it must be establish optical power budget. The loss
budget specifies the minimum and maximum amount of loss margin that can be tolerated
in between the OLT and ONU.
Passive Optical Networks (PON) is a point-to multipoint optical network with
no active elements in the signals path from source to destination. Figure
1 shows the FTTH PON basic architecture. The only interior elements used
in such networks are passive combiners, couplers and splitters (Yeh
et al., 2008). In order to share the available total bandwidth, a bidirectional
1xN splitter is used in a P2MP PON. The splitter is a bidirectional broadband
branching device that has one input port and multiple output ports. The input
optical signal is divided among the output ports, allowing multiple users to
share a single optical fiber and consequently share the available bandwidth
of that fiber.
||FTTH-PON which used the Intermediate Split Structure (ISS)
In the upstream direction, optical signals from a number of ONTs are combined
into the single fiber. In the upstream direction, data frames from any ONU will
only reach the OLT, not another ONUs due to directional properties of a passive
combiner such as optical splitter. In this case, in the upstream the behavior
of EPON is similar to that P2P architecture (Kramer and
Pesavento, 2002). The passive coupler is positioned not more than 1 km from
residential customers, in order to minimize fiber usage. Each customer receives
a dedicated short optical fiber but shares the long distribution trunk fiber
(Png et al., 2005).
A typical EPON architecture is composed of one Optical Line Termination (OLT)
and Multiple Optical Network Unit (ONU) connected to the OLT via a passive optical
splitter (Chen et al., 2004). EPON vendors are
focusing initially on developing Fiber To The Business (FTTB) and Fiber To The
Curb (FTTC) solution, with the long term objective of realizing a full service
Fiber To The Home (FTTH) solution for delivering data, video and voice over
a single platform (Vacca, 2007).
EYE DIAGRAM PARAMETER
This study focused on the eye diagram parameter analysis in the FTTH-PON network
during the working and failure condition. Four parameters will be concentrated
which are BER, eye opening, maximum Q factor and jitter. In the case of failure
condition, four restoration schemes are activated to ensure the signal flow
continuously without perturbation. In this research, the eye diagram that shown
in Fig. 2 is used to analyse the working and failure condition
in the FTTH-PON network. The analysis is done based on four parameters in order
to determine the quality of received signal. The parameters are Eye height,
Maximum Q Factor, Jitter and Eye area.
||Eye diagram parameters which determine the quality of receive
Eye height: It refers to the distance from the base to the peak of the
eye measured in voltage. The allowed minimum height value of the eye diagram
is inversely proportional to the photosensitive sensitivity. High photosensitive
sensitivity could assess data at low height value of eye diagram whereas low
photosensitive sensitivity requires high eye diagram height value for data assessment
(for example, at photosensitive sensitivity-22.8 dBm, the eye diagram height
value is 5.5 μV and at sensitivity-18 dBm, the allowable eye diagram height
value is 10 μV). If there is any fixation on the photosensitive sensitivity
value, the increase in data transmission rate will give similar eye diagram
Maximum Q factor: The maximum Q factor refers to the quality of the
produced eye diagram to be analyzed. This value is fix and equal for all photosensitivity
values for various data transmission rate. The allowable maximum Q factor is
6 in current communication system to obtain BER value equivalent to 1x10-9.
Jitter: Deterministic jitter refers to the shifting which occurs on
derivative and embedded time of the received signal and the value is measured
according to the duration at the crossing point. It is measured in UI (unit
interval) unit and the allowable maximum value is 0.2 UI. Normally the jitters
are generated and give pronounce impact on the data transmission system of more
than 1 Gbps.
Eye area: Eye area refers to the distance between levels of bit 0 and
bit 1 and the distance between right and left embedded derivatives time crossing.
The parameters are applied in mask technique to evaluate the quality of the
received signal. The width of this area is important in the process of differentiating
bit 1 and bit 0 as well as the sequence of the first and second bits. The more
wide the eye areas is, the bigger the received data quality is and thus, the
process of signal sampling becomes easier.
OVERVIEW OF PROPOSED RESTORATION SCHEME FTTH-EPON
In this study, we concentrated on four types of failure mechanism. For every
type of protection mechanism, we employ the dedicated protection and shared
protection. If traffic prioritization is implemented, high priority traffic
is transmitted on the primary path whereas the best effort traffic is diffused
on the backup path. In case the primary path breaks, the high priority traffic
is transferred to the backup path. The failure protection and recovery of services
needs the following actions: When the breakdown occurs, then it must be detected
and the information about the failure has to be propagated to the nodes triggering
protection switching actions. For switching the service from a failed working
path to a backup path, then the backup path has to be set up. Thus, a suitable
route with sufficient resources has to be found for the backup path means that
a pre-established backup path has to be disjointing from the working path. Resources
need to be allocated to the backup path. Finally, the service has to be switched
over to the backup path. The described actions may take place at different points
For dedicated path protection, a working path and an end to end backup path
is established and resources are assigned to it at connection set up time. There
is a difference between 1:1 and 1+1 protection. In 1+1 protection, the client
signal is transported simultaneously on both working path and backup-path, whereas
in 1:l protection, the client signal is switched over from the working to the
backup path after the occurrence of a failure. In this case, the backup resources
can be used for pre-emptive services in the absence of a failure on the working
However, from a resource perspective, both mechanisms are the same. Hence,
we will not make a distinction between the two mechanisms and refer to them
as dedicated path protection. In our case study we assume the working path to
be link disjoint from the backup path which ensures to protect against any single
link failure. In shared protection, backup paths are pie-calculated and sufficient
backup resources are preplanned. However, backup paths are assigned to specific
protection paths only after a failure occurrence. Sufficient in this context
means that the backup resources are dimensioned so that service recovery from
any single link failure is guaranteed. Many different ways of sharing resources
and many approaches for dimensioning and optimizing the shared backup resources
are possible. It is unrealistic to plan and optimize the backup resources for
all services of a dynamic network in a common step, but it has to be done sequentially
at service request by service request. We establish the working path on the
shortest available route, and calculate a link-disjoint backup path, which reuses
as much backup resources of previously established services as possible. Only
those backup resources of other services are used which are not subject to the
link failures on the established working path. As needed, additional backup
capacity is reserved. The concept provides 100% protection guarantee in case
of single link failures. The nodes compute the working and the backup path based
on information about whether a wavelength on a link is used as backup or working
resource, or whether it is free. Using this method, little additional backup
resources are assigned.
Working condition A: In Fig. 3, the grey arrow shows
the normal network condition when there is no failure occurs in the working
line. As the working line is in a good condition, the optical signal could be
sent through it and the protection line is not being used.
Failure condition B: Figure 4 shows the failure is
detected in working line, protection mechanism will be activated and convert
the optic signal direction to the protection line. The arrow shows the protection
mechanism as dedicated protection.
Condition failure C: Figure 5 shows the shared protection
scheme when breakdown occurs in both line in working line and protection line.
Shared protection scheme will be activated and optic signal will convert the
route to neighbor line protection as depicted with black arrow.
Condition failure D: Figure 6 shows the breakdown
occurs in both lines and shared protection scheme will be activated and the
signal will be routed to the next neighbour protection line. However when the
failure also occurs in the neighbour working line, then the mechanism design
protocol will give priority for dedicated protection and the signal will be
routed to the next neighbour line protection which is in normal condition as
in light grey arrow. The dark grey arrow represent the protection mechanism
which convert the optic signal to the second neighbour protection line as the
first neighbour protection used for dedicated protection (black arrow).
|| Protection mechanisms in ideal condition
||Breakdown occurred at working line and signal is diverted
to the protection line
||Breakdown occurred at working line and protection line. Signal
is diverted to the neighbour protection line
||Breakdown occurred at working line, protection line and working
RESULTS AND DISCUSSION
The FTTH based network design was modeled and simulated using the Optisystem
CAD program by Optiwave System, Inc. The two optical fibers were connected between
the transmitter and 1:8 bidirectional splitter (18 km) using a bidirectional
optical fiber also the other one was linked between splitter and ONUs (2 km)
by using Single Mode Fiber (SMF). In the downstream direction, at the OLT, two
wavelength channels which are 1550 and 1480 nm are multiplexed and transmitted
in optical fiber (18 km) to the bidirectional splitter.
|| Simulation parameters
In the upstream direction the 1310 nm wavelength was transmitted. Simulation
aims to verify the network system feasibility and investigate the system performance
of the proposed protection route mechanism based EPON architecture. In this
simulation we proposed FTTH-EPON design will have 8 ONUs. A transmission distance
between OLT and ONU is 20 km. The 1480 and 1550 nm downstream signals and 1310
nm upstream signal have 1.25 Gb sec-1 direct modulation in the test
access network. And the output powers of 1480 and 1310 nm lasers are 0 dBm.
Moreover, the power budget of the architecture as follows. In normal condition,
1480 and 1550 nm signals will traverse one circulator bidirectional (1dB), bidirectional
optical splitter (3dB) and about 20 km Single Mode Fiber (SMF) (5 dB), one multiplexer
(0.5 dB), one demultiplexer (0.5 dB) and two numbers of optical switch (2.4dB)
thus, the total loss budget is about 12.4 dB. The sensitivity of optical receiver,
which is used in our test system, is nearly to -34 dBm. The Bit Error Rate (BER)
performances are measured by a 1.25 Gb sec-1 Non-Return-to-Zero (NRZ)
Pseudo Random Binary Sequence (PRBS) with a pattern length of 231-1
for the downstream traffic between the OLT and 8 ONUs. The specifications of
components in this simulation model are tabulated in Table 1.
Our results were obtained by observing bit error rates, eye diagrams, optical
power levels and dispersion levels.
Eye diagrams show parametric information about the signal which is the effects
deriving from physics such as system bandwidth health. It will not show protocol
or logical problems. If a logic 1 is healthy on the eye, this does not reveal
the fact that the system meant to send a zero. However, if the physics of the
system mean that a logic one becomes so distorted while passing through the
system that the receiver at the far end mistakes it for a zero, this should
be shown in a good eye diagram.
||Observed eye diagrams for (a) 1550 nm downstream at working
condition A, (b) 1550 nm downstream at condition failure B, (c) 1550 nm
wavelength at condition failure C and (d) 1550 nm wavelength at condition
|| Signal parameter from the eye diagram
Figure 7 a-d show the eye diagram for downstream
wavelength. Clear opening eye diagram were observed for working condition A
rather than failure condition D. Failure at condition D gives the highest value
of dynamic range since in the failure condition, the protection route mechanism
uses eight numbers of optical switches to perform the protection and restoration
scheme to the network. The results in the form of eye diagrams from which various
signal parameters can be calculated. Table 2 gives the value
of calculated parameter such as Q factor, eye opening, jitter and BER value
for all condition types. There is a reduction in jitter, eye opening, BER and
Q factor for failure condition when it use many protection route to implement
the restoration scheme as in condition D.
For every number of optical switch, the insertion loss is considered equal
to 1.2 dB. In this simulation, the values can be accepted and above the minimum
requirement which is 6 (~ BER = 1x10-9) was achieved. For every type
of protection mechanism, we employ the dedicated protection and shared protection.
According to the four failure conditions which is condition A, condition B,
condition C and condition D the protection route will involve in two, four,
six and eight numbers of optical switch, respectively.
Figure 8a and b show the eye diagram for
(a) the downstream data at 1.25 Gb sec-1 in -25 dBm sensitivity and
(b) -34 dBm sensitivity. From the result achieved, the simulation model was
simulated in -25 dbM and-34 dBm sensitivity. Clear opening was observed at-34
dBm sensitivity. The receiver performance is characterized by measuring the
BER as a function of the average optical power received. The average optical
power corresponding to a BER of 10-9 is a measure of receiver sensitivity.
By using the optimization in receiver sensitivity, we found that receiver sensitivity
managed to be adjusted in -25 dBm sensitivity using goal attainment type of
optimization which is commonly used for parameter extraction.
|| Eye diagrams observed from 1550 nm wavelength in (a) -25
dBm sensitivity (b) -34 dBm sensitivity
||Effect of BER on the launch of the enhanced up to 3 dBm. BER
obtained decreased exponentially when the launch power increases
Then the thermal noise of a PIN extracted to obtain receiver sensitivity that
equal to-25 dBm. For this simulation we set the receiver sensitivity to-34 dBm
by using the SPO optimization. From the result achieved, the proposed architecture
design in all condition can only been used effectively in-34 dBm sensitivity,
since the receiver sensitivity of -25 dBm is not manage to provide good system
In optical receivers, a receiver is said to be more sensitive if it achieves
the same performance with less optical power incident on it. The performance
criterion for digital receivers is governed by Bit-Error Rate (BER), which is
defined as probability of incorrect identification of a bit by the decision
circuit of the receiver. A commonly used criterion for digital receivers requires
the BER to be below 1x10-9 (Max Q Factor ~6).
Simulation platform has been built to simulate the performance of each state
of failure is available. The objective of this study is to examine the effects
of restoration on the activation of the system BER performance.
|| Q factor values of a time frame for each of the faults in
Input power was changed from -3 dBm to 3 dBm and BER observed. Observations
made on the BER profile formed on the restoration schemes. It was found that
the signal attenuation can be seen clearly when the damage is increasing. The
slight narrowing of the signal obtained across the normal state without failure.
Figure 9 shows the effect of BER on the launch of the enhanced
up to 3 dBm. BER obtained decreased exponentially when the launch power increases.
Is possible if the power is reduced to -2 dBm launched on the damage level 4,
the bit rate will be increased up to 1e-08. If the signal is too weak, then
the data signal will be difficult to separate the noise and causes the error
received are increasing.
Figure 10 shows the effect of changing the maximum Q factor
with a bit of time in which quality is obtained with a high Q factor in the
normal state. The situation has resulted in damage to the four factors that
diminished the quality and turn into shrinking the Q factor of 6.4.
The survivability of FTTH network is necessary to provide seamless services
and ensure network reliability. Single failure in the line connected will activate
the dedicated protection while shared protection is activated when both fiber
(working and standby fiber) are breakdown. The BER characteristics were measured
at 1.25 Gbps and no degradation was observed, as confirmed by a comparison of
these simulation results with those obtained from systems without restoration
element. The simulation model and the results were presented to convince proposed
protection scheme. Survivability in network system will provide the protection
and restoration architectures and it continued services in the presence of failures.
The survivability will add redundant capacity, detect faults and automatically
re-route traffic around the failure. So the mechanism of restoration for the
system was designed to meet the network specification.
This research project had been carried out in the Computer and Network Security
Research Group laboratory under the Dept. of Electrical, Electronics and System,
Universiti Kebangsaan Malaysia (UKM). This project was started on 1st July of
2010 and was fully sponsored by Universiti Kebangsaan Malaysia (UKM) with the
grant number UKM-PTS-082-2010. The duration of the project is about 24 months
and it was categorized under ICT cluster. All authors and co-authors are contributed
in this project.