Optics and Photonics Journal, 2012, 2, 8-12

http://dx.doi.org/10.4236/opj.2012.21002 Published Online March 2012 (http://www.SciRP.org/journal/opj)

Behavioral Variations of Gain and NF Owing to

Configurations and Pumping Powers
Belloui Bouzid
Hafr Al-Batin Community College (HBCC), King Fahd University of Petroleum and Minerals (KFUPM), Dhahran, Saudi Arabia
Email: Bouzid@hbcc.edu.sa

Received November 29, 2011; revised January 5, 2012; accepted January 12, 2012

ABSTRACT
Six configurations are proposed in this paper to explore the gain and noise figure (NF) variations under the pumping
power effect. I propose a new investigation of gain and NF at different EDFA configurations. Configurations such as
SPSS, DPSS, DPSSF, TPDS, TPDSF, and QPDSF are designed, investigated and compared. A continuous progress of
gain values is observed from SPSS to QPDSF, and a change of NF values related to configurations is recorded. The NF
variations show different behaviors at different configurations. High gain of 59.49 dB and low NF value of 4.22 dB are
recorded for the QPDSF configuration and low gain and low NF are recorded for the SPSS configuration.

Keywords: Double Pass; Single Pass; Erbium Doped Fiber; Configuration; Pumping Power

1. Introduction milestone role in this crucial era.

The lack of entire description of behavioral study of
EDFA is a crucial milestone in optical communication
gain and NF at different configurations in the published
systems and wide growing internet. The progressive de-
papers will affect the fast growing of optical amplifier
velopment of EDFA since its early stage at the end of
and reduce its effectiveness. Most of the published pa-
eightieth is showing a continuous progress at the two
pers are focusing on NF and efficiency [1], gain and gain
levels of EDFA knowledge, theoretical and experimental. flattening [2-5], and optimization of pumping power [6]
EDFA is an important part in the long haul optical fiber without any research papers which, focuses on the de-
communication, and it is considered as another interesting scription of the complete behavioral variations and trends
research topic for laser phenomena of spontaneous and of EDFA gain and NF at different configurations.
stimulated emission, where interaction of matter-light- All factors affecting the EDFA are needed to be elabo-
matter still in its early stage. rated and investigated to enhance the amplification out-
Fiber to the home will guide the future changes of puts. Focusing on improvement of gain and noise figure,
communication in the near future. Laser with the stimu- the devised and optimized configuration is to establish a
lated and spontaneous emission are considered to be the wide and flat EDFA gain [2-5], and to discover an effi-
main factor in the next quantum revolution. Let us think cient EDFA at all levels. The obligation for an efficient
how terabits transmission can be carried out without fiber EDFA is to fulfill the practical need of communication,
optics or fiber amplifier? How the narrow band of elec- with high capacity and high speed.
tronics can be handling the wide broad band of commu- The EDFA description at different configurations is
nication systems? At the beneath of the laser phenomena rarely found in the published papers [7], and the use of
there are many promising future development and new the entire physical phenomena to describe, illustrate, and
discoveries that can be conditionally achieved. Optical interpret the variations of gain and NF at different con-
amplifier and lasers are used nearly in all the wide spec- figurations is highly required to understand their trends at
trum of science such as medicine, military, education and large-scale.
manufacturing. In this paper general descriptions with illustrations and
Research in optical communication systems is growing analysis are performed based on the studies of six differ-
daily. In addition, the obligation for efficient systems to ent configurations: single pass single stage (SPSS), dou-
fulfill the practical need of communication for high ca- ble pass single stage (DPSS), double pass single stage
pacity and high speed is extremely demanded due to the with filter (DPSSF), triple pass double stage (TPDS),
fast growing of high speed and high capacity transmis- triple pass double stage with filter (TPDSF) and quadru-
sion. Highly efficient and stable EDFAs is playing a ple pass double stage with filter (QPDSF) [8].

Copyright © 2012 SciRes. OPJ

 B. BOUZID 9

2. Experiment Setup and Discussion of power, and the wavelength division multiplexer (WDM)
Results is to merge both 980 nm pump and 1550 nm signal in
The used erbium-doped fiber in this experiment is char- EDF. The signal will be reacted with stimulated emission
acterized by: NA of 0.27 cutoff wavelength of 840 nm, and amplified spontaneous emission (ASE) where at the
peak absorption at the signal 1527 nm wavelength of 6 output the ASE and the stimulated signal will be dis-
dB/m, erbium concentration of 440 ppm, and Er3+ core played. The filter will eliminate almost all ASE; the OSA
doped in silica/germania. The New port, tunable band displays the amplified signal with small portion of ASE
pass filter (TBF) is mechanically tuned with a pass-band at the bottom of the signal.
of 1 nm, an insertion loss of 1.5 dB at the tuned wave- The six configurations were shown in Figure 1(a)
length, and a tuning range limited to 45 nm from 1520 to SPSS: single pass single stage, Figure 1(b) DPSS: dou- ble
1565 nm. The Tunable Laser Source (TLS) is a continu- pass single stage, Figure 1(c) DPSSF: double pass single
ous wavelength source of the 1550 nm input signal stage with filter, Figure 1(d) TPDS: triple pass double

Figure 1. Experimental configurations of EDFA: (a) single pass single stage (SPSS); (b) double pass single stage (DPSS); (c)
double pass single stage with filter (DPSSF); (d) triple pass double stage (TPDS); (e) triple pass double stage with filter
(TPDSF) and (f) quadruple pass double stage with filter (QPDSF). TBF: tunable bandpass filter, CIR: circulator, EDF:
erbium-doped Fiber, LD: laser diode, and WDM: wavelength division multiplexing, INPUT: tunable laser source, and
OUTPUT: optical spectrum analyzer.

Copyright © 2012 SciRes. OPJ

10 B. BOUZID

stage, Figure 1(d) TPDSF: triple pass double stage with at the changing of configurations from SPSS, DPSS,
filter, and QPDSF: quadruple pass double stage with fil- DPSSF, TPDSF and QPDSF. Except for the TPDS with-
ter. The difference between these configurations is owing out filter, which shows a lower gain compared to DPSSF
to the additions of TBF and the second stage which can and TPDSF.
be single pass or double pass. The circulators are used as It can be seen clearly, by following the variations of
loop back where port1 and 3 are spliced and the TBF is gain values versus configurations, at lower pump power
incorporated between these ports to suppress and elimi- of 10 mW, the gain is varied at different configurations
nate the unwanted ASE. The Key role of TBF in this where a shift between 9.65 and 45 dB of the gain values
continuous increase of gain is impressive and crucial is recorded for SPSS and QPDSF respectively. At higher
where stimulated emission will strongly amplify the signal. pumping power of 90 mW the gain is shifted owing to
Due the large number of configurations used in these configurations change from 20.04 to 59.49 dB. All these
experiments, we will describe only the signal in the results are at 1550 nm input signal power and –50 dBm
QPDSF configuration. The configuration in Figure 1(f) except the SPSS. This good result shows clearly the var-
showed that each turn consistently gave a signal attenua- ied configurations, the filter, and the double pass impact
tion of 12 dB, this loss is due to three circulators, two on the gain values. So, with the change of configuration
TBF filters, and two WDMs. Thus, the amplified signal from SPSS to QPDSF, the gain is increased to 45 dB at
will propagate through the CIR1 from port1 to port2 then low pumping power. The gain gap between the SPSS and
travel through EDF1; the signal will be affected by the QPDSF reach to 39.45 dB at high pump power.
first amplification from EDF1, through port2 into port3 Go in details for the shown results in the figure. The
of CIR2, passing through the first TBF1 filter into port1 SPSS records the lowest gain where the QPDSF records
and back to port2 to be amplified during the second pass the highest one. At this level of explanation, the effect of
by EDF1 into port2 of CIR1, and therefore will propa- the configurations type is very crucial where the gain gap
gate again in the second stage through EDF2, CIR3, and between the SPSS and QPDSF reach approximately
TBF2 for the third and fourth passes. The output signal 35.83 dB at 10 mW pump power. It is evident that the
power was displayed through the OSA from port4 of effect of the pumping power and configuration with filter
CIR1. Traveling from port1 to port4 of CIR1, the signal are effective for the gain enhancement. It is also observed,
will be affected by four amplifications during the four that the DPSSF has a higher gain compared with TPDS
passes, or, as we mentioned, the quadruple pass double where single stage has higher gain compared to double
stage with filter configuration [8]. stage. I think, the role of filter in the design is certainly
Figure 2 shows gain on dB versus configurations and crucial and principal in this reversed phenomenon. In this
pumping power. The input signal power is at 1550 nm case it is recorded that adding single stage single pass to
wavelength and –50 dBm only the SPSS. From the 3D the double pass configuration will reduce the gain.
graph, all the six configurations show an increase of gain Figure 3 shows experimental NF versus configura-

Figure 2. Experimental gain versus configurations and pumping power at 1550 nm wavelength at –50 dBm input signal
power. SPSS: single pass single stage, DPSS: double pass single stage, DPSSF: double pass single stage with filter, TPDS:
triple pass double stage, TPDSF: triple pass double stage with filter, and QPDSF: quadruple pass double stage with filter.

Copyright © 2012 SciRes. OPJ

 B. BOUZID 11

Figure 3. Experimental NF versus configurations and pumping power at 1550 nm wavelength at −50 dBm input signal power.
SPSS: single pass single stage, DPSS: double pass single stage, DPSSF: double pass single stage with filter, TPDS: triple pass
double stage, TPDSF: triple pass double stage with filter, and QPDSF: quadruple pass double stage with filter.

Table 1. Gain and NF variations due to configuration. shown in this paper, that the configuration structure and
the filter are crucial in the variation of the gain and NF.
Configurations Gain NF
In particular, high gain can be generated simply with
SPSS LOW LOWEST simple modification related to the design structure.
QPDSF records 59.49 dB gain and 4.22 dB NF. This
DPSS LOW HIGH
result can be increased higher with fifth or sixth passes
DPSSF HIGH LOW and higher pump power. Table 1 shows the description
TPDS LOW HIGH and the comparison of gain and NF at different configu-
rations.
TPDSF HIGH HIGH

QPDSF HIGHEST LOW 3. Conclusion

The configurations SPSS, DPSS, DPSSF, TPDS, TPDSF,
tions and pumping power at input signal power of 1550 and QPDSF were exposed to the same conditions of input
nm wavelengths and −50 dBm except SPSS. It is evident signal power wavelength and different pumping power.
that the NF is decreased at specific configurations and The full investigation of gain and NF shows high gain of
increased at others. The 3D graph shows clearly that the 59.49 dB and low NF of 4.22 dB values of QPDSF con-
NF is decreased at three configurations. It is also evident figuration. Gain and NF were profoundly described and
that the NF is bellow 5 dB for the SPSS, DPSSF, and clearly investigated using both 3D and 2D graphs display.
QPDSF configurations, and above 5 dB for the DPSS, A continuous increasing of gain values was demonstrated.
TPDS, and TPDSF configurations. The TPDS and TPDSF The gain progress was recorded from SPSS, DPSS,
show the highest NF compared with the other configura- DPSSF, TPDS, TPDSF, to QPDSF configurations. A
tions. variation of NF related to different configurations was
It can be noted the effect of adding the single pass
proved and demonstrated. The TBF positioned in the
stage to the double pass on increasing of noise figure. In
between the circulator ports, configurations structure, and
the double stage double pass the TBF filter is affecting
pumping power are a principal factors for gain and NF
the QPDSF and the DPSSF and enhance their NF. The
control.
configuration SPSS records the lowest NF with the low-
est gain, QPDSF records the highest gain. This can be
4. Acknowledgments
related to the multi factors that have been included such
as: double pass, double stage, and the filter positioned The author wishes to acknowledge UPM, MMU (Malay-
between the two ports of the circulator. It is clearly sia) and HBCC/KFUPM (Saudi Arabia) for their support

Copyright © 2012 SciRes. OPJ

12 B. BOUZID

in providing the various facilities utilized in the presenta- Tellurite Fiber Amplifier for WDM Signals in the 1581 -
tion of this paper. 1616 nm Wavelength Region,” Electronics Letters, Vol.
36, No. 36, 2000, pp. 621-622.
[5] C. Yang, “Design and Simulation of Gain-Flattened Ultra
REFERENCES Wideband Fiber Amplifiers Covering S-, C-, and L-
[1] A. C. Çokrak and A. Altuncu, “Gain and Noise Figure bands,” Ph. D Thesis, The University of North Carolina at
Performance of Erbium-Doped Fiber Amplifiers,” Jour- Charlotte, Charlotte, 2003.
nal of Electrical & Electronics Engineering, Vol. 4, No. 2, [6] M. A. Mahdi, K. A. Khairi, B. Bouzid and M. K. Abdul-
2004, pp. 1111-1122. lah “Optimum Pumping Scheme of Dual-Stage Triple-
[2] B. Bouzid, B. M. Ali and M. K. Abdullah. “A High Gain Pass Erbium-Doped Fiber Amplifier,” IEEE Photonics
EDFA Design Using Double Pass Amplification with a Technology Letters, Vol. 16, No. 2, 2004, p. 419.
Band-Pass Filter,” Photonics Technology Letters, Vol. 15, doi:10.1109/LPT.2003.821059
No. 9, 2003, pp. 1195-1197. [7] A. Sellami, K. Al-Khateeb and B. Belloui, “The Influence
doi:10.1109/LPT.2003.814901 of EDFA’s Configuration on the Behavioral Trends of
[3] Y. B. Lu and P. L. Chu, “Gain Flattening by Using Dual- Gain,” International Conference on Computer and Man-
Core Fiber in Erbium-Doped Fiber Amplifier,” IEEE agement, Kuala Lumpur, 9-11 May 2006, pp. 853-856.
Photonics Technology Letters, Vol. 12, No. 12, 2000, pp. [8] B. Bouzid, “High-Gain and Low-Noise-Figure Erbium-
1616-1617 . Doped Fiber Amplifier Employing Dual Stage Quadruple
[4] A. Mori, T. Sakamoto, K. Shikano, K. Kobayashi, K. Pass Technique,” Optical Review, Vol. 17, No. 3, 2010,
Hoshino and M. Shimizu, “Gain Flattened Er3+-Doped pp. 100-102.

Copyright © 2012 SciRes. OPJ