International Journal of Scientific & Engineering Research, Volume 5, Issue 7, July-2014 718

ISSN 2229-5518

Performance Enhancement of Radio over Multimode Fiber System using Fiber Bragg Grating for Micro and Pico Cell Applications

Shuvodip Das, Ebad Zahir

Abstract— Radio over Fiber (RoF) is a promising technology for short range transmission applications within multimode optical fiber. Typically, the RoF link employs a single mode fiber. But the signal power at the remote antenna become small due to the power loss in the electrical to optical and optical to electrical conversion process. Coupling efficiency of an electrical to optical converter can be improved with multimode fiber. But multimode fiber suf- fers from dispersion. Therefore, the paper proposes a simplified and efficient Radio over Multimode Fiber (RoMMF) system that can reduce the effect of dispersion by employing Fiber Bragg Grating (FBG). The performance analysis based on maximum Q-factor, minimum BER, threshold, eye height and jitter are extensively investigated. In addition, the effect of fiber length on these performance metrics is presented. Finally, a FBG based RoMMF system is proposed by providing a side-by-side comparison between the proposed and traditional system for micro and pico cell applications.

Index Terms— Bit Error Rate (BER), Dispersion, Eye height, Fiber Bragg Grating (FBG), Jitter, Optisystem, Q-factor, Radio over Multimode Fiber

(RoMMF)

1 INTRODUCTION

—————————— ——————————
ADIO over fiber (RoF) refers to a technology whereby light is modulated by a radio signal and transmitted over an optical fiber link to facilitate wireless access, such as
3G and WiFi communication simultaneously from the same antenna. When radio signals are carried over multimode fiber cable then it’s called Radio over Multimode Fiber (RoMMF) system. This system consists of Central Station (CS) and Base Station (BS) or Radio Access Point (RAP) connected by a mul- timode optical fiber link or network as shown in Fig. 1. Usual- ly, RoF system utilizes Single Mode Fiber (SMF). [1], [2], [3]

Fig. 1. RoF System Concept [3]

The choice of optical fiber in RoF system depends on applica- tions. For short range applications, such as micro cell (less than 2 kilometer) and pico cell (about 200 meter or less), mul- timode fiber is preferred for its cost effectiveness and copling efficiency. Moreover, the signal power at the remote antenna is

————————————————

Shuvodip Das is currently pursuing a masters degree program in Electrical and Electronic Engineering at American International University - Bangladesh. E-mail: shuvo@primeuniversity.edu.bd

Ebad Zahir, Assistant Professor, Department of Electrical and Electronic

Engineering at American International Unversity - Bangladesh. E-mail:

ebad.zahir@aiub.edu
very small due to significant power loss in the electrical to optical and optical to electrical conversion process. However, coupling efficiency of an electrical to optical convertor can be improved using multimode fiber (MMF) as the core of MMF is more than five times greater in diameter than that of a SMF. The larger core enables low loss connection and facilitates simple fiber-to-fiber or fiber-to-transceiver alignment and con- sequently is best suited for premises of the range of micro and pico cells. [4] Moreover, deployment of multimode fiber leads to reduction in cost of the link. But Step-Index MMF (SIMMF) experiences Intermodal dispersion that degrades the perfor- mance of the system and quality of the received signals. On the other hand, Graded-Index Multimode Fiber (GIMMF) con- taining parabolic refractive index profile decreases modal dis- persion. Dispersion is the main parameter which needs to be compensated in order to provide high level of reliability of service. Fiber Bragg Grating (FBG) is one of the most widely used element to compensate dispersion. FBG is a periodic per- turbation of the effective refractive index in the core of an op- tical fiber that generates a wavelength specific dielectric mir- ror. So, FBG can be used as an inline optical filter to block cer- tain wavelengths. [5], [7]

Fig. 2. Working principle of FBG [6]

IJSER © 2014 http://www.ijser.org

International Journal of Scientific & Engineering Research, Volume 5, Issue 7, July-2014 719

ISSN 2229-5518


The simulation used ideal dispersion compensation FBG with
user-defined group delay. The transfer function of the filter,

H ( f ) = e jf ( f )

(1)
Where, f is the frequency dependence phase of the filter.
Group delay depends on wavelength as

λ2 df τ (λ ) =

(5)
Eye pattern or diagram is used to visualize how the wave- forms used to send multiple bits of data can potentially lead to errors in the interpretation of those bits. Vertical eye opening indicates the amount of difference in signal level that is pre-
sent to indicate the difference between one bit and zero bit.

2πc dλ

Where, c is the speed of light.

Phase, f = 2πcτ (λ )

1 dλ

(2)
The bigger the difference the easier it is to discriminate be- tween one and zero. Whereas, horizontal eye opening indi-
cates the amount of jitter present in the signal. An open eye

λ (3) [8]

In our work, we recommended the RoF link that employs a
Parabolic Index Multimode Fiber (PIMMF) and FBG as inline
optical filter to increase coupling efficiency and reduce disper-
sion. Simulation results from Optisystem 12 have been includ- ed to show the comparative performance evaluation of our proposed system containing PIMMF and FBG and traditional system without using FBG. Parameters like maximum Q- factor, minimum BER, eye height, threshold and jitter with respect to varying fiber length have been considered. Simula- tion results show that the proposed system exhibits acceptable performance, considering Q-factor, BER, eye pattern and jitter at maximum of 0.75 kilometer which makes the system suita- ble for simple, cost effective micro and pico cell applications.

2 PERFORMANCE MEASURES

Characterization of an optical transmission link which is one of the main criterions for the effective modeling of RoF system depends on the proper choice of performance metrics. Per- formance metrics should present a precise determination of system’s limitation and measurement to improve the perfor- mance of the system. The most widely used performance measures are the Q-factor, BER, eye opening and jitter.
Q-factor is helpful as an intuitive Figure of Merit (FoM) that is
directly tied to the BER. BER can be improved by either a) in- creasing the difference between the high and low levels in the numerator of the Q-factor or b) decreasing the noise terms in the denominator of the Q-factor.

Q = VH VL

pattern corresponds to minimal signal distortion. Distortion of the signal waveform due to intersymbol interference and noise appears as closure of the eye diagram. [8]
Moreover, understanding of jitter is important because all dig-
ital circuits require at least one, and often several clocks for processing and handling data. As gate count and processing speeds increase, chipset designers must take into account fac- tors such as propagation delay, skew, rise/fall times, etc to en- sure adequate margins for proper operation. Jitter is one such factor. It adds uncertainty to the exact timing of an external reference clock.
Total jitter (T) is the combination of random jitter (R) and de- terministic jitter (D):

T = Dpeak-to-peak + 2× n×Rrms (6)

In which the value of n is based on the bit error rate (BER) re- quired of the link. [10]

3 METHODOLOGY AND SIMULATION SCHEMATIC

One of the main objectives of this paper is to simulate and model a FBG based RoMMF system which leads to simple and cost effective system implementation. Fig. 3 depicts the block diagram of the proposed RoMMF system.

Fig. 3. Block diagram of our proposed FBG based

RoMMF system

σ L + σ H

(4)
VS is the voltage sent by the transmitter and if we assume that
Vs can take on one of the two voltage levels, VH and VL . σL
and σH are the standard deviations of the noise. [9]
On the other hand, BER gives the upper limit for the signal because some degradation occurs at the receiver end. The bit error probability, Pe is the expectation value of the BER. The BER can be considered as an approximate estimate of the bit error probability. In a noisy channel, the BER is often ex- pressed as a function of the normalized carrier-to-noise ratio (Eb /N0 ), (energy per bit to noise power spectral density ratio), or Es /N0 (energy per modulation symbol to noise spectral den- sity).
In Fig. 4, Central System (CS) is the spatial optical transmitter
subsystem consisting of pseudo-random binary sequence gen- erator (PRBS), coding/modulation block (NRZ and RZ), laser, AM modulator, DC bias and multimode generator. The central wavelength of laser is 850nm. In our proposed system, we have used wavelength which falls under A-Band and facili- tates the use of lower cost Vertical-Cavity Surface-Emitting Laser (VCSEL). Optical channel that comprising spatial con- nector attaches signals with transverse mode profiles. To re- duce the effect of dispersion, SIMMF is replaced with PIMMF having core diameter of 62.5 µm. After that ideal dispersion compensation FBG of a center wavelength of 850 nm is used as

IJSER © 2014 http://www.ijser.org

International Journal of Scientific & Engineering Research, Volume 5, Issue 7, July-2014 720

ISSN 2229-5518

filter to block unwanted wavelengths. Radio Access Point (RAP) is the spatial optical receiver consisting of spatial aper- ture and optical receiver, illustrated in Fig. 5.

Fig. 4. Spatial optical transmitter subsystem

Fig. 5. Spatial optical receiver subsystem

1 Gbps data from PRBS is converted into NRZ and RZ signal sequence and the last stage of the optical transmission facili- tates the user to select between an External Modulated Laser (EML) or a Directly Modulated Laser (DML). After that the RF modulated optical signal is fed to the multimode generator that attaches transverse mode profiles to the input signal. It also converts a single mode signal into a multimode signal based on the user defined power distribution. Power ratio ar- ray parameter is fixed to “1 2 3 4” that means the multimode generator generates four spatial modes per polarization (LG00 , LG 22 , LG 03 and LG 13 ). Four spatial modes then propagate through PIMMF. Parabolic refractive index helps to reduce modal dispersion. Afterward, ideal dispersion FBG having dispersion of -160 ps/nm filters out the undesired wavelengths and at output we get a 850nm concentrated output. Then the spatial aperture of spatial optical receiver subsystem couples the optical signal from the multimode fiber to the photodetec- tor. Later, the signal is fed to the optical receiver built using a PIN photodetector, a Bessel filter and a 3R regenerator. PIN photodetector converts optical signal into electrical signal. Whereas, Bessel filter having center wavelength of 850nm at- tenuates signals with wavelengths other than 850nm. Finally,
3R regenerator regenerates an electrical signal. It generates the original bit sequence and a modulated electrical signal to be used for BER analysis. FBG based RoMMF system is modeled using simulation software Optisystem 12. Fig. 6 shows the
simulation schematic drawn in Optisystem 12 window using various in-built blocks.

Fig. 6. Simulation schematic of proposed FBG based

RoMMF system

4 SIMULATION RESULTS AND ANALYSIS

Proposed RoF system was successfully modeled and simulat- ed using Optisystem 12 to extract simulation results. In this specific design, we have employed four types of visualizers, spatial visualizer, optical time domain visualizer, optical spec- trum analyzer and BER analyzer.
Fig. 7 and Fig. 8 present the mode and weighted mode profile
of optical field at the output of the a) spatial optical transmit- ter and b) PIMMF respectively.

(a) (b)

Fig. 7. Spatial visualizer displays the transverse mode at the output of the

(a) spatial optical transmitter and (b) PIMMF.


(a) (b)

Fig. 8. Spatial visualizer displays the weighted mode profile at the output of the (a) spatial optical transmitter and (b) PIMMF.

Fig. 9 illustrated the time domain representation of signal at the output of (a) spatial optical transmitter and (b) PIMMF after 0.5 km fiber length. Distortion in Fig. 9 (b) resembles the effect of attenuation and dispersion.

IJSER © 2014 http://www.ijser.org

International Journal of Scientific & Engineering Research, Volume 5, Issue 7, July-2014 721

ISSN 2229-5518


(a) (b)

Fig. 12. 3D BER graph at (a) 0.5 km and (b) 0.75 km

(a) (b)

Fig. 9. Optical time domain visualizer displays the time domain signal at the output of the (a) spatial optical transmitter and (b) PIMMF.

The consequence of inserting FBG is shown in Fig. 10. This figure represents the optical spectrum before and after FBG for
0.5 km fiber length. Fig. 10 (b) shows narrower spectrum con-
centrated on 850 nm than that of Fig. 10 (a). FBG filters out
unwanted spectrum and make the resultant spectrum narrow- er.

(a) (b)

Fig. 10. Optical spectrum visualizer displays the optical spectrum a) before and b) after FBG

Fig. 11 shows the BER pattern and Q-factor after 0.5 km and
0.75 km. In the Fig. 11 (b), for 0.75 km eye width and eye clo- sure reduced due to jitter effect and intersymbol interference.

(a) (b)

Fig. 11. Eye diagram showing BER pattern and Q-factor after (a) 0.50 km and (b) 0.75 km

Fig. 13 shows the total jitter of an eye diagram, measured at
the eye cross point, as the difference between the time values of marks A and B. Fig. 13 (a), (b) and (c) show the jitter for 0.5 km, 0.75 km and 1km.

Fig. 13. Jitter for (a) 0.5km (b) 0.75km and (c) 1km

Table I: Q-factor, Minimum BER, Eye Height and Threshold for Different values of Fiber Length for RoMMF system without FBG.

IJSER © 2014 http://www.ijser.org

International Journal of Scientific & Engineering Research, Volume 5, Issue 7, July-2014 722

ISSN 2229-5518

Table II: Q-factor, Minimum BER, Eye Height and Threshold for Different values of Fiber Length for RoMMF system with FBG.

Based on the simulation data extracted from Table I and Table II, Fig. 14 (a), (b), (c) and (d) have been drawn. The results shown in Fig. 14, depicts the performance of RoMMF with and without FBG considering maximum Q-factor, minimum BER, eye height and threshold with varying fiber length.

(a) (b)

(c) (d)

Fig. 14. Graph of (a) maximum Q-factor (b) minimum BER

(c) Eye height and (d) Threshold against varying fiber length for RoMMF

system with and without FBG

5 CONCLUSION

In this paper, we have proposed a FBG based RoMMF system and simulated the transmission of 1 Gbps data carried over
0.25 km to 1.25 km PIMMF at wavelength of 850 nm. The sim-
ulation results shown in Figure 7-13 depicted the performance of FBG based RoMMF system for different fiber lengths. Pa- rameter metrics such as, maximum Q-factor, minimum BER, eye height, threshold and jitter have been considered. Figure
14 showed the performance comparison between RoMMF sys-
tem with and without FBG. Figure 8 showed the weighted transverse mode profile at the output of spatial transmitter and PIMMF. Figure 9 (b) showed small amount of deviation in
the time domain signal at the output of PIMMF due to attenu- ation and dispersion. Figure 10 (b) illustrated the optical spec- trum after FBG and demonstrated the effectiveness of using FBG as it made the linewidth narrower by filtering out unde- sired wavelengths. After been transmitted for 0.5 and 0.75 km, received signal accumulated some noise as shown in the eye diagram, 3D BER graph and jitter in Figure11, 12 and 13. Eye diagram for 0.25 km and 0.5 km having wider vertical and horizontal eye opening corresponds to minimal signal distor- tion. Simulation results showed that with the increase of fiber length Q-factor, eye height and threshold decrease and mini- mum BER and jitter increases in most instances. Comparison between RoMMF system with and without FBG has been made based on the performance metrics, such as, Q-factor, minimum BER, eye height and threshold. System with FBG showed slightly better performance than that of system with- out using FBG. Our proposed FBG based RoMMF system dis- played acceptable performance for maximum fiber length of
0.75 km. Acceptable eye diagram and low BER were achieved
which implies the better performance of the system for micro and mainly for pico cell applications.

REFERENCES

[1] R. Karthikeyan, S. Prakasam, PhD., “A Survey on Radio over Fiber (RoF) for Wireless Broadband Access Technologies”, International Journal of Computer Applications (0975 – 8887) Volume 64– No.12, pp. 14-19, February 2013.
[2] Ahmad Said Chahine, Uche A. K Okonkwo and Razali Ngah, “Study the Per- formance of OFDM Radio over Fiber for Wireless Communication Systems”,

2008 IEEE INTERNATIONAL RF AND MICROWAVE CONFERENCE PRO- CEEDINGS, December 2-4, 2008, Kuala Lumpur, Malaysia

[3] Shuvodip Das and Ebad Zahir, “Modeling and Performance Analysis of RoF System for Home Area Network with Different Line Coding Schemes Using Optisystem”, International Journal of Multidisciplinary Sciences and Engineer- ing, Vol. 5, Issue 6, pp – 1-8, June 2014.
[4] Roland Yuen, Xavier N. Fernando, Sridhar Krishnan, “Radio over Multimode Fiber for Wireless Access”, IEEE Canadian Conference on Electrical and Com- puter Engineering, May 2014.
[5] Kishore Bhowmik, Md. Maruf Ahmed and Md. Abdul Momin, “Reduction of Dispersion in Optical Fiber Communication by Fiber Bragg Grating and Opti- cal Phase Conjugation Techniques”, International Journal of Mobile Network Communications and Telematics, Vol. 2, No. 3, pp – 49-58, June 2012.
[6] "Fiber Bragg Grating Principle", http://www.fbgs.com/technology/fbg- principle/, Last visited on: 15th July 2014.
[7] Ojuswini Arora, Dr. Amit Kumar Garg, “Impact of Fiber Bragg Grating as Dispersion Compensator on the Receiver Characteristics”, Global Journal of Researches in Engineering Electrical and Electronics Engineering, Vol. 11, Issue

7, December 2011.

[8] “Optical Signal-to-Noise Ratio and the Q- Factor in Fiber-Optic Communica- tion Systems”, Matrix Integrated, Application Note: H FAN-9.0.2
[9] Amanjot Kaur, Jasbir Singh, “Performance Evaluation of Digital Modulation Techniques in a WCDMA-based Radio-over-Fiber Communication System”, International Journal of Advanced Research in Computer Science and Electron- ics Engineering, Volume 1, Issue 4, pp. 10-14, June 2012.
[10] Xuelian Ma, Lu Liu, and Junxiong Tang, “Timing jitter measurement of trans- mitted laser pulse relative to the reference using type II second harmonic gen- eration in two nonlinear crystals”, Optics Express, Vol. 17, Issue 21, pp. 19102-

19112 (2009)

IJSER © 2014 http://www.ijser.org