Docstoc

A Novel Passive Optical Add_Drop Multiplexer _OADM_ for WDM

Document Sample
A Novel Passive Optical Add_Drop Multiplexer _OADM_ for WDM Powered By Docstoc
					A Novel Passive Optical Add/Drop Multiplexer (OADM) for WDM Systems

                 Ruzita Abu Bakar and Nejim Abd Ali Al-Asedi
                               TEMAN, MIMOS Bhd
                Technology Park, 57000, Kuala Lumpur, Malaysia
               Tel: +60 3 89965000 ext. 4244 Fax: +60 3 89960529
                       Email: ruzita@mimos.my, nejim@mimos.my

                Mohd Khazani Abdullah and Harith Ahmad
     Telekom Photonic Research Center (TMPRC), University Malaya, Jalan
                       Lembah Pantai, Kuala Lumpur

                           Borhanuddin Mohd Ali
 Faculty of Engineering, University Putra Malaysia (UPM), Serdang, Selangor


                                          Abstract

The introduction of IP over Wavelength Division Multiplexer (WDM) network is one way of
enabling the creation of a super-high speed network in a very economical way.

Optical Add Drop Multiplexer (OADM) is one of the devices for WDM system in enabling the
Telecommunication Companies to reroute traffic signals while preserving the integrity and
flexibility in handling various signals of other channels. Moreover, it offers ranges of services
to customers who require the multiplexing capability to add and/or drop channels.

The work described in this paper was carried in the Telekom Photonic Research Center
(TMPRC) of University Malaya as part of a postgraduate study with University Putra Malaysia
(UPM). The TMPRC is focusing on research and development works in fiber optics devices
for future telecommunication technologies.

This paper present the original research works carried out on the development of the 1st
Malaysian novel passive OADM to be used in WDM systems. The work is Patent Pending.


Keywords: OADM, WDM, Drop channel, Add Channel, Directional Coupler, Bragg grating,
Isolator, Insertion loss, Crosstalk, Return Loss



1.      Introduction

The origin of optical networks is linked to WDM which provide additional
capacity on existing fibers. The optical network is proposed to provide end-to-
end services completely in the optical domain, without having to convert the
signal to the electrical domain during transit [1]. Transmitting IP directly over
WDM has become a reality and is able to support bit-rates of OC-192. As we
can clearly see, it holds the key to the bandwidth surplus and opens the
frontier of high-speed telecommunication in the new century.

With the introduction of WDM systems, new roles emerged for the use of fibre
optics in telecommunication system. WDM is currently gaining popular
supports all over the world as it offers many advantages over traditional TDM
systems that transmit only a single channel [2]. The advantages include a


                                               1
much higher channel capacity, less fibres to be installed, and channel
management flexibility. To implement and realize the WDM technology for
fibre-optic communication systems, several new optical components are also
being researched and developed quite extensively [4]. Among the important
components in the WDM system are multiplexer and demultiplexer, tunable
optical filters, wavelength routers, optical cross-connects, wavelength
converters, WDM transmitter and receiver and Optical Add/drop
multiplexers(OADM).


2.    Introduction to OADM

OADM is the optical sub-system that facilitates the evolution of the single
wavelength point-to-point optical network to the WDM networks. The OADM
serves as the entry point to the optical layer in many other aspects. The
practical utilization of the fiber bandwidth is achieved by being able to
selectively remove and reinsert individual channels, without having to
regenerate all of the WDM channels. An OADM is characterized in terms of
the total number of inputs in the drop and add channels. The system
maintains each connection as sequential ports and performs manipulations on
them.

OADM enables carriers to reroute traffic to different geographic areas without
requiring extensive termination equipment [3]. It gives great flexibility to
carriers who offer a range of services to their customers and require the
multiplexing capability to add and/or drop channels in both directions [4] as
shown in Figure 1. These can also be done at an optical amplifier site without
interrupting the channels. Moreover, OADM eliminates the need for back to
back terminals, thus reducing the carrier network equipment cost as well as
enabling the carrier to exploit fibre bandwidth, improve network reliability and
reduce the cost of providing broadband services [5].


      Input                                 OADM
                                                      Output
      (1-n)                                         (,2-n,x )



                                     Drop     Add
                                     1       x


Figure 1: Block diagram of a general OADM

OADM has the ability to add and drop certain wavelength channels from a
fibre while keeping other channels undisturbed is an example of optical
routing capability [6]. OADM are needed for WDM networks in which one or
more channels need to be dropped or added while preserving the integrity of
other channels as shown in Figure 2. This provides a great flexibility in
handling the various signals.



                                       2
                                                                                  Data 1
     Data 1
                                                 1
     Data 2           MUX                                                         Data 2
                                                                  DE-
                                          ADD/                    MUX
     Data 3                               DROP

                                                      EDFA
                             EDFA                                                 Data 3
                                                      Amplifier
                             Amplifier
     Data 4                                    d
                                                                                  Data 4

Figure 2: A WDM fibre optic link

OADM resembles demultiplexer-multiplexer pair since its operation requires
demultiplexing of the input WDM signal, changing the data content of one or
more specific wavelength channels and then multiplexing the entire signal
back again [7]. Figure 3 shows a generic add/drop Multiplexer. In this system
it is possible to amplify the WDM signal and equalize the channel powers at
the OADM since each channel can be individually controlled [8].


                 Demultiplexer                             Multiplexer
                                   1 Optical switch
      WDM                                                                WDM
      Input                        2                                    Output



                                   n


Figure 3: Generic OADM using optical switches [8].

If a single channel needs to be demultiplexed and no active control of
individual channels is required, one can use a multiport device that sends a
single channel to one port while all other channels are transferred to another
port, thus avoiding the need for demultiplexing all channels. Such devices are
often called add/drop filters since they filter out a specific channel without
affecting the WDM signal. If only a small portion of the channel power is
filtered out, such device acts as an optical tap as it leaves the contents of the
WDM signal intact.


3.     Novel Passive OADM (NP-OADM) Design and Development

The NP_OADM work was carried in the Telekom Photonic Research Center
(TMPRC) of University Malaya as part of a postgraduate study with University
Putra Malaysia (UPM). The design present the original research works carried
out as the 1st Malaysian prototype of NP_OADM.



                                         3
The proposed design is a passive OADM, that add/drop a single channel.
Passive components are non-powered devices such as lenses, isolators,
micro-collimators and optical attenuators which work with active or powered
devices such as lasers and laser-light modulators, to optimise the quality of an
optical signal [5]. In this design the passive components used are two 3 dB
directional couplers (tapper), one isolator and one Bragg grating as shown in
Figure 4. The input port was connected to the main line from which a plurality
of channels (1 - n) propagates into the device. The channels generate
through the isolator and the first directional coupler. The isolator was used to
block the back reflection from the Bragg grating.


     Input Port (1-n)
     Pin                Isolator


                                                                          Path 1
                                    Tapper A
                                                 1-n     Bragg
                                                          grating
   Path 2
                                                               IL3       ,2-n
                                                     d
                                                                          Tapper B   2-n ,x
               Drop Port Pd
               d
                                                    x                                Output
                                                                                      Output Port
                                                                                      Port Po
                                      Add port
                                                                Path 3                   Pout
                                      Pa x

Figure 4:The schematic diagram of NP-OADM with a single channel
dropped

The channels were then split according to the 3 dB coupling ratio by the
tapper 1, while the required wavelength (d) is reflected by the Bragg grating
and sent through another power splitting before dropping out through drop
port. A tapper provides flexibility in the design since the tapper’s coupling
ratios can easily be changed. The rest of the wavelengths (2 - n) are sent
through the tapper B according to its coupling ratio back to the main line
through the output port. Simultaneously, the required channel (x) is
extracted by the added port propagating to tapper B to the main line. Finally,
the rest of the channels (2 - n) combine with the added channel (x) and
execute out of the device through the output port.

The system is configured in all-fiber devices; therefore, there is no conversion
required from electrical form to optical signal. As a result the system has little
access losses and interference. This design is very flexible since the power
splitting can be controlled by the coupler ratio using variable coupler [9].
Furthermore, the development of the device does not require the expertise of
highly technical personnel; thus, reduces the production cost. Another
advantage of the device is the drop and adds channel can easily be
customised to any channels requested by the customer by inserting the
suitable Bragg grating. The component cost is also low since it requires one
Bragg grating only, while the cost of the two directional coupler is not so much
of a factor (directional coupler are becoming common items).



                                         4
4.    NP-OADM Characterisation and Analysis

The design was then functionally tested. The characterization of each system
will then follow the emphasis of the design in relation to insertion loss,
crosstalk and return loss of the systems.

4.1   Functional Test

Functional test is performed in order to certify that the system function as
specified. In other words, the system can drop the required channel and add
channel simultaneously from the main trunk. This means that there should be
an increase of power measured at the drop port when the transmitted
wavelength match the Bragg grating reflection wavelength (drop channel). At
the same time the power measured at the output port should drop.

The interchannel effect of two channels applied simultaneously at different
input power has been studied. The two channels used are at 1538.75 nm
(Channel 1) and 1541.80 nm (Channel 2) with the input power of –28.27 dBm
and –13.43 dBm, respectively as shown in Figure 5. The reflected channel is
obtained at Channel 2.




                                            Channel 2
                                            1541.80 nm
                    Channel 1
                    1538.75 nm




Figure 5: The two input channels used for the functional test

Figure 6 shows the power measured at the dropped port. Channel 2 gives -
23.73 dBm implying that 59% power was extracted from the main trunk. The
rest of the power is transmitted through the output port as shown in Figure 7.
The power measured for Channel 2 at the output port was –71.10 dBm that is
48.73 dBm lower than the power measured at the drop port. The percentage
of power out is also influenced by the quality of Bragg grating used in the
system. Figure 6 and Figure 7 confirm that this system could be successfully
used for WDM system.




                                      5
                                                 Channel 2

                         Channel 1




       Figure 6: Power measured at the drop port Channel 1 at –56.45
       dBm and Channel 2 at –23.73 dBm




                       Channel 1

                                                 Channel 2



Figure 7: Power measured at the output port with Channel 1 at –37.84
dBm and Channel 2 at –71.10 dBm


4.2.   Insertion Loss test

Insertion loss of the system is measured from the test using equation 1 as
listed in appendix I. Two sets of test have been studied; insertion loss for
single power input and double powers input. In a single power input test, a
single laser source was varied from 1535 nm to 1545 nm signal wavelength
with input power of about 8 dBm. The insertion lost of reflected channel which
measured at the output port give maximum value at 27.03 dB, whereas the
insertion loss measured at the drop port give minimum value at 8.28 dB as
shown in Figure 8 [10].

For insertion loss of double input powers measured at drop port as shown in
Figure 6, the insertion loss for ILPath 2 at Channel 2 was 10.30 dB, which is
2.02 dB higher than insertion loss using single input channel and the insertion
loss for Channel 1 was 28.18 dB. This is due to the crosstalk between the
channels.




                                      6
          Insertion loss (dBm)   40
                                 30                                          Path 2
                                 20
                                                                             Path 1
                                 10
                                  0
                                  1530      1535        1540          1545
                                          Wavelength (nm )



Figure 8: Insertion loss versus wavelength for Path 1 and Path 2 for a
single input channel test

The maximum insertion loss of the system is calculated theoretically based on
the schematic design Figure 4. Refer to section B of the appendix for the
formula. The theoretical value for Insertion loss for drop port is 13.1 dB.

From this value it shows that the insertion lost measured theoretically for the
system is higher than tested. For example, a single input power gave 8.28 dB
insertion loss which is 4.82 dB lower and for a double power input test gave
10.30 dB which is 2.8 dB lower than theoretical. These results testify that this
design perform better than expected.


4.3   Crosstalk test

Crosstalk on NP-OADM is due to the intrinsic Bragg grating and the Fresnel
reflections of all channels.

To evaluate the impact of the crosstalk on the system performance, both
paths (Path 1 and Path 2) crosstalk were measured as shown in section C of
the appendix. The test result for the crosstalk measured at the drop port is
tabulated in Table 1. The average crosstalk measured for was 19.90 dBm.

Test                             Pd            PT            Xpath1
1st                              -21.15        -41.00        19.85
2nd                              -21.25        -41.12        19.87
3rd                              -21.14        -41.07        19.93
4th                              -21.18        -41.08        19.90
5th                              -21.22        -41.13        19.91
Average                                                      19.90

Table 1: The average value of crosstalk for path 1




                                                        7
The test result for the crosstalk measured at the output port is tabulated in
Table 2. The average crosstalk measured for Path 2 was 20.32 dBm.

Test         Pout T           Pd            Xpath2
1st          -20.35           0             20.35
2nd          -20.30           0             19.87
3rd          -20.34           0             19.93
4th          -20.32           0             19.90
5th          -20.31           0             19.91
Average                                     20.32

Table 2: The average value of crosstalk for path 2

The minimum requirement for crosstalk on telephony application is 20 dB and
the international standard product specification for OADM is 25 dB.
Therefore, this design is performing within acceptable range.


D.    Return lost

Common sources of reflection are glass-air interfaces at open fibre ends,
mechanical splices, crack, poorly mated connectors because of mud and dirt,
and sometimes overpolished connectors. Reflections not only cause light
returning back to the source, but also cause loss of optical power and
reflection by the Bragg grating.

Four tests had been conducted as discussed in section D of the appendix and
tabulated as shown in Table 3. The average of return loss was 4.92 dBm.
This value shows that return loss of the design is so minimal and within
acceptable range.

Test            Pin                Po                   Return lost
1st             -4.92              0                    4.92
2nd             -4.94              0                    4.94
3rd             -4.90              0                    4.90
4th             -4.91              0                    4.91
Average                                                 4.92

Table 3: The average value of return loss. Pin is the power input from the
laser source and Po is the power detected/reflected from the NP-OADM


E.    Add Channel Test

The next test that has been performed is the added channel test. Four input
channels used were 1537 nm(Channel 1), 1538.53 nm (Channel 2), 1541.90
nm (Channel 3) and 1564 nm (Channel 4) with the power input of –1.20 dBm,
-1.22 dBm, 1.25 dBm and 0.09 dBm respectively. Channel 1,2 and 3 were
injected through the input port and at the same time Channel 4 is injected
through the add port. The dropped channel is Channel 2.


                                        8
Figure 9 shows the power measured at the drop port for Channel 1, 2 , 3 and
4 which were –8.50 dBm, –8.50 dBm, –8.0 dBm and –38.3 dBm, respectively.
The insertion loss for the Path 2 measured at drop port are as follows:
Channel 1 at 7.30 dB, Channel 2 at 7.28 dB, Channel 3 at 9.22 dB and
Channel 4 at 37.91 dB.

The difference on the reflected channel is 0.05 dB compared with the previous
test which indicates that the added channel did not interfere much with the
drop channel of the system.




                                              Channel 2
                                              (–8.50 dBM)

              Channel 1
              (–8.50 dBM)
                                           Channel 3
                                           (-8.0 dBm)

                                                        Channel 4
                                                        (-38.3 dBm)




Figure 9: Power spectrum for NP-OADM at Channel 1,2,3 and 4
measured at add port

The added channel has been successfully transmitted through the output port
as shown in Figure 10. The power at Channel 3 detected was –19.93 dBm
which is higher than the power measured at the drop port. It indicates that
added functionality is within the specification. The insertion loss for Path 1
measured at the output port is as follows: Channel 1 is 2.43 dB, Channel 2 is
1.8 dB, Channel 3 is 60.55 dB, and Channel 4 is 20.02 dB. It shows here that
the insertion lost for added channel is quite high which is 20.02 dB this is due
to the crosstalk and the insertion lost of tapper 2 of the system.



                                      Channel 2                 Channel 4
                Channel 1
                                      (–3.65 dBm)               (-19.93 dBm)
                (–3.00 dBM)



                                               Channel 3
                                               (-59.3 dBm)




Figure 10: Power spectrum of NP-OADM for Channel 1,2,3 and 4 measure at the
output port



                                       9
5.      Conclusion

This paper discussed the development NP-OADM and various
characterization test of the system. The proposed NP-OADM was described
in details. Issues of characterization of the system such as insertion loss,
crosstalk, and return loss were also discussed.

The test results testify the system can be successfully used for WDM system
by terminating the intended channel and executing the rest of the channels.
From the characterization test, the system is found to perform within the
acceptable range. The insertion lost was found at 8.28 dB for single channel
input test and 10.30 dB for two channels input test. Whereas, the crosstalk
measured at the drop port was 19.90 dB and at the output port was 20.32 dB.
The return loss was 4.93 dB which is so minimal.

The system can be improved in term of performance by using circulator, or
combiner but with slightly higher cost. Improvement of the loss factor and the
multi channels drop was suggested for further studies.



REFERENCES


[1]     Srinivasan      Seetharaman,      “IP    over      DWDM”            ftp://ftp.netlab.ohio-
state.edu/pub/jain/courses/cis788-99/ip_dwdm/index.html , pp 1

[2] Global Information, Inc. “Optical Add/Drop Multiplexer Market Brief, http://www.giii.co.jp ,
1998

[3] Grace Murphy, http://www.broadband-guide.com/1w/news/nes1986.html, NEC Trials 32-
channel DWDM with Carrier, Lightwave Xtra, News, January 1989
[4]
  Wavelength division multiplexing increases the capacity of transport networks, News, Issue
12, May 1997, http://www.tmo.hp.com/tmo/tcnews/9705/12tncov.html,

[5]“Migration to All Optical Network”, http://www.usa.alcatel.com , 1998

[6] George Ishikawa and Hiroki Oi, “WDM Systems – Photonic Networks”, Fujitsu
Laboratories Ltd, 1998

[7] F. Shehadeh, R. S. Vodhanel, M. Krain, C. Gibbons, R.E. Wagner, and M. Ali, IEEE
Photon. Technol. Lett. 7, USA, 1995, pp1075

[8] Giles, Randy and Spector, Magaly, "The Wavelength Add/Drop Multiplexer for Lightwave
Communication Networks", Bell Labs Technical Journal, Vol 4, No 1, Jan-Mar 1999, pp.207 -
229.

[9] , Ruzita Abu Bakar, Mohamad Khazani Abdullah, Fazlee Isnin, Burhanuddin Ali and Harith
Ahmad “Tapper Ratio Dependency Of A Low-Cost Passive OADM System”, Proceeding Lislo,
99, UKM, Bangi

[10] Ruzita Abu Bakar, Mohd Khazani Abdulllah, , Borhanuddin Mohd Ali, Malay Mukherjee,
Nejim Al-Asedi, Harith Ahmad, “Loss And Cross-Talk Studies Of A Novel Optical Add/Drop
                                         st    nd
Multiplexer (OADM)”, NCTT98 Conference, 1 – 2 December, 1998, IDEAL, UPM, Serdang




                                              10
Appendix 1
A       Test measurement of Insertion Loss test

Insertion loss of the system is measured from the test as follow

                 IL Path 1 = 10 log (Pin /Pout)                                                [1]
                                       n
                 ILPath 2 = 10 log      Pin i                                                 [2]
                                       i=1
                                                      n
                                                        Pd       i
                                                      i=1


where ILPath1 is the insertion loss measured from input port to the output port and ILPath 2 is
the insertion loss measured from input port to the dropped port. Whereas, P in is the power
input used throughout the test, Pout is the total output power measured from output port and
Pd is the power drop measured at the drop port.

B.      Theoretical measurement of Insertion Loss

The maximum insertion loss of the system is calculated theoretically based on the schematic
design Figure 4 as follow

                          IL Path 2 = ILiso + 2(ILT1) + ILBragg                                [3]

where ILiso is the insertion loss for Isolator, ILT1 is the insertion loss for directional coupler 1
and ILBragg is the insertion loss of Bragg grating. The value of the insertion lost is taken from
the characterization of components, ILiso is 0.75 dB, ILT1 is 3.9 dB and and ILBragg is 4.55 dB.

C.      Test Measurement of Crosstalk

To evaluate the impact of the crosstalk on the system performance, both paths (Path 1 and
Path 2) crosstalk were measured. The crosstalk measured at Path 1 is as follows

                          Xpath1 = 10 log           Pd                                         [4]

                                                     PT
                                                               d = 0

where Xpath1 is the crosstalk measured at the Path 1 and

                          Pd  i = 0                                                          [5]

Pd is the power measured at the drop port when the reflected wavelength (d) is on and the
other signal (i) is off.

                          PT  d = 0                                                          [6]

Whereas PT is power measured at the output port when the d were off and the i were on.

The crosstalk measured at Path 2 is as follow

                          Xpath2     = 10 log            Pout T                                [7]

                                                          Pd
                                                                      out = 0




                                                   11
where Xpath2 is the crosstalk measured at Path 2 and

                Pout T  d = 0                                                    [8]

Pout T is the power measured at output port when the reflected wavelength (d) is off and other
wavelength (i) is on.

                Pd    out    i = 0                                               [9]

Whereas Pd out is the power measured at output port when the d is on and i is off.

D.      Test measurement of Return Lost

The return lost of the system is measured from the test as follow

                Return lost        = 10 log (Pin / Po)                             [10]

where Pin is the power input from the laser source and Po is the power detected/reflected from
the OADM as shown in Figure A1.


             Laser
             source
                                  Pin                                Pa
                                                                            OADM
                                                         Tapper

             Detector
                                        Po




Figure A1: Measurement for return loss of OADM system consist of laser source,
detector, tapper and OADM




                                                   12

				
DOCUMENT INFO