Journal of Engineering and Technology

: 2012  |  Volume : 2  |  Issue : 1  |  Page : 19--23

Error Vector Magnitude Analysis of Radio-over-fiber Systems Based on Single Side Band Modulation

Shelly Singla, Sandeep Arya 
 Guru Jambeshwar University of Science and Technology, Hisar, Haryana, India

Correspondence Address:
Shelly Singla
Guru Jambeshwar University of Science and Technology, Hisar, Haryana


In order to meet the increasing demand for wireless connectivity and large coverage, radio-over-fiber (RoF) infrastructure has been suggested as a cost-effective solution for the provisioning of bandwidth in small cell size. In the present work, performance analysis has been done for radio over systems based on optical single side band technique. Error vector magnitude (EVM) has also been quantified for RoF link in the presence of dispersion and phase noise incorporating dual electrode Mach-Zehnder modulator (DE-MZM). It is found that the laser phase noise is dominant over long distances.

How to cite this article:
Singla S, Arya S. Error Vector Magnitude Analysis of Radio-over-fiber Systems Based on Single Side Band Modulation.J Eng Technol 2012;2:19-23

How to cite this URL:
Singla S, Arya S. Error Vector Magnitude Analysis of Radio-over-fiber Systems Based on Single Side Band Modulation. J Eng Technol [serial online] 2012 [cited 2020 Sep 23 ];2:19-23
Available from:

Full Text

 1. Introduction

Facing explosive demands of high channel capacity, wider service coverage and broadband radio-over-fiber (RoF) system entail a technology that can meet those requirements in the coming future. Hence, RoF technology is the most promising solution for enhancing the capacity and mobility as well as lessening costs of the base stations (BSs) where most of signal processing such as RF generation, coding, multiplexing and modulation can be done at the central station (CS). However, the performance of RoF systems depends on the method used to generate the optically modulated radio frequency (RF) signal, power degradation due to fiber chromatic dispersion, nonlinearity due to an optical power level, and phase noises from a laser and an RF oscillator. There are two techniques to generate the optically modulated RF signal: direct and external modulation. The direct modulation scheme is simple but suffers from a laser-frequency chirp effect, and this chirp effect results in severe degradation of the system performance. However, this can be eliminated by using the external-modulation scheme instead of the direct modulation scheme [1] . Although the external-modulation scheme is employed, the conventional optical double sideband (ODSB) signal can degrade the received RF signal power due to fiber chromatic dispersion drastically. For overcoming the power degradation, an optical single sideband (OSSB) signal, generated by using a phase shifter and a dual-electrode (DE) Mach-Zehnder modulator (MZM), is employed [2] . In addition to these two effects, the nonlinearity of an optical fiber can give a large penalty on the long-haul transmission and multi channel system using a high-power signal. For the high-power transmission, the nonlinear effect should be managed by utilizing a modulation format [3] . Kitayama et al., [4] analyzed the system performance for an ODSB signal including laser phase noise and suggested how to compensate the differential delay by using a dispersion- compensating fiber (DCF). He focused on how to compensate fiber chromatic dispersion for the ODSB signal experimentally and analytically rather than analyze the effect of the phase noise on the performance in detail. Barry and Lee [5] and Salz [6] analyzed the performance of coherent optical systems with laser phase noise by utilizing a Wiener process, since coherent detection provides better sensitivity than that of direct detection, while direct detection has a simple structure. Gallion and Debarge [7] and Tkach [8] used an autocorrelation function and a PSD function for evaluating the effect of the laser line width and fiber chromatic dispersion on the system performance. In [9] , the CNR penalty due to the laser line width is negligible in a narrow laser line width and small differential delay (100ps) while the CNR penalty is quite large in a broad laser line width and large differential delay. Sharma et al., [10] analyzed the impact of spectral width of laser over intensity noise introduced inside the fiber incorporating higher order dispersion parameters and showed that intensity noise can be reduced by reducing the laser line width to kHz range in long-haul communication systems. A technique for measuring residual single sideband (SSB) microwave phase noise, added by an externally modulated fiber-optic link, was reported in [11] . A model for calculating additive phase noise in direct modulation optical links was presented in [12] . Chromatic dispersion effects on the phase noise of optical millimeter-wave systems were investigated in [13] and [14] for direct and remote heterodyne detection. Performance comparison of systems for various modulation formats have been reported by [15] . Number of papers has been published covering the modelling, analysis of performance characteristics and measurement of phase noise in optical links within various contexts. In this work, we have studied by simulation that effect of laser-spectral width in a single-tone OSSB-RoF transmission system incorporating DEMZM modulator and derived EVM for the system, an important performance characteristic in RoF systems. EVM is a measure of errors between the measured symbols and expected symbols. The use of EVM as a performance metric is limited to radio frequency engineering to infer reception the performance at the receiver.

 2. Theoretical Modeling and Analysis

An OSSB signal is generated by using dual electrode MZM and a phase shifter. A RF signal from an oscillator is split by a power splitter and a 90° phase shifter. This RF signal is optically modulated by the laser diode (LD) with an MZM. The optically modulated signal is transmitted to the photo diode (PD) and the photocurrent corresponding to the transmitted RF signal is extracted by the filter. First, the optical signals from the optical source, laser diode and the RF oscillator are modeled as:


Where, A d and Vo define amplitudes from the optical source and the RF oscillator signal, ωd and ωo define angular frequencies of the signals from the LD and the RF oscillator, and Φd (t) and Φo (t) are phase-noise processes. The OSSB signal generated using Dual electrode MZM is modeled in equation (3).


After the transmission of signal over L km fiber, the signal can be represented as equation (4) & in this equation Ladd denotes an additional loss in the optical link, αfiber is the SSMF loss, Lfiber is the transmission distance of the SSMF, and τ0 and τ+ define group delays for a center angular frequency of ωd and an upper sideband frequency of ωd + ωo . φ1 and φ2 are phase-shift parameters for specific frequencies due to the fiber chromatic dispersion.


The photocurrent i(t) can be obtained as [INSIDE:1]

Where η defines the responsibility of the PD and [INSIDE:2] is the square-law detection.

To evaluate the SNR, we utilize the autocorrelation function and the PSD of the photocurrent.


The autocorrelation function R 1 (τ) is obtained as

R 1 (τ) = (i(t).i(t+ τ))

Now we will evaluate PSD function which is Fourier transform of RI(τ)

S1(f) = f 〈 R1 (τ)ρ

The received RF carrier Power P rcd is approximately represented as follows


Signal-to-noise ratio (SNR) can be used to predict the performance of the system. The SNR induced by the differential delay from the fiber chromatic dispersion and the line widths from the laser and the RF oscillator is found as:


 3. Error Vector Magnitude (EVM) Analysis

EVM measurements are often performed on vector signal analyzers (VSAs), real-time analyzers or other instruments that capture a time record and internally perform a Fast Fourier Transform (FFT) to enable frequency domain analysis. Signals are down converted before EVM calculations are made [16] . The EVM is defined as the root-mean-square (RMS) value of the difference between a collection of measured symbols and ideal symbols. These differences are averaged over a given, typically large number of symbols and are often shown as a percent of the average power per symbols of the constellation.


Where η = responsivity, Ad1 = constant related to the laser light amplitude and the losses in fiber, MZM and the joint and splices given by [INSIDE:3] J= Bessel function of 1 st kind, of order n and α1 = normalized RF voltage given by [INSIDE:4] where Ad1 is the amplitude of laser light, L MZM is the lose in the MZM, L add is the factor accounting for the additional loss in the fiber, αfiber is the loss in the fiber and Lfiber is the length of fiber. Vrf is the input RF voltage and Vπ is the MZM switching voltage, p is the ratio of the power required for a particular filter used to the total carrier power. This parameter incorporates the effect of the bandwidth of the filter being used and No is the additive white Gaussian noise power spectral density. The parameters 2γLD = 2π∆VLD and 2γRF = 2π∆VRF , define the angular full-linewidth at half maximum (FWHM) of the lorentzian shape for the laser and the RF oscillator and 2γt = 2π∆VLD + π∆VRF gives the total line width. τ = τ ± τ o is the differential delay due to the fiber chromatic dispersion and is given by [INSIDE:5]

Where, D is the fiber chromatic dispersion parameter, Lfiber is the fiber length, fRF is the RF frequency and c is the speed of light.

The first parameter is the photodiode responsivity η. For most of the photo diodes its values is between 0.6 and 0.8. Taking the value of η as 0.7 [11] . Now the second constant is [INSIDE:6].

Here LMZM is the loss of the DE-MZM. Now considering the MZM as an integrated waveguide power splitter and combiner, its value can be assumed to be negligible. L addis the additional loss caused by the fiber components such as the splices, joints, etc. Its value for a 10-km fiber link can be taken as approximately 3 dB and is varied accordingly for various length of fiber. αfiber is the loss per km of the fiber and is around 0.2 dB/km for SSMF. Lfiber is the length of the fiber and is equal to 10 km for this case. α is the modulation index of the MZM and is equal to α=Vrf / Vπ. Now taking V rf = 1 mV and Vπ=2.2V, we obtain α=0.00045 then the modulation index is given as απ = 0.0014. It gives equal to 1 approximately. From above all, the value of is calculated as 0.1342. N0 is the power spectral density of the AWGN for very low noise case, it can be taken as 10-11. Now α1 depends upon the first harmonic of the photo detector and the fundamental component. So the value of α1 is 0.001. Thus all the constants terms of EVM are evaluated and its derived expression is then used to study the effects of the laser line width and dispersion on the EVM of the RoF system.

 4. Simulation Parameters


 5. Result and Discussion

The impact of laser line width, ΔVLD is described in [Figure 1] and [Figure 2] and results are calculated with RF oscillator line width of 0.1Hz, 0.8Hz and 1Hz for OSSB-RoF transmission system with laser line width varying from 100 to 700 MHz as a function of laser phase through SSMF fiber of different optical links (10-30 km) and first-order dispersion of 17 ps/nm km. In order to investigate the effect of the line width, we set the P-value as 0.5 in any situation, and it means that the same type of filter, such as the half-power bandwidth, is utilized for all cases. It can be seen from [Figure 1] that EVM has an increasing trend with increase in laser line width over various RF oscillator line widths of 0.1 Hz, 0.8 Hz and 1 Hz, as the line width of RF oscillator is usually less than 1 Hz, for a 10-km fiber. An increase of 2 dB is observed as the laser line width is swept from 100 to 700 MHz and from [Figure 2] it can be been seen that laser noise is dominant over long distances. The results are calculated for 10-30 km OSSB-RoF transmission system as it requires less bandwidth than DSSB-RoF system and is tolerable for power degradation due to a chromatic fiber-dispersion, through a standard single-mode fiber carried. The EVM due to the laser line width increases dramatically over a specific distance. For 30 km exponential increment is observed giving an increment of about 7 dB in EVM while the rate of increment is lesser at 20 km and 10 km. Therefore, the laser line width should be selected carefully in a long-haul transmission since the large differential delay and large laser line width cause serious performance degradation.{Figure 1}{Figure 2}

 6. Conclusions

EVM performance analysis of single side band-based RoF system employing MZM has been carried out theoretically and numerical simulation has also been done. It is observed that EVM increases exponentially with increase in laser line width with respect to transmission distance and a gradual increment exist with increase in RF oscillator line width and laser line width.


1G. H. Smith, and D. Novak, "Overcoming chromatic-dispersion effects in fiber- wireless systems incorporating external modulators", IEEE Trans Microwave Theory Tech, vol. 45, no. 8, pp. 1410, 1997.
2J. Leibrich, "CF-RZ-DPSK for suppression of XPM on dispersion managed long-haul optical WDM transmission on standard single-mode fiber", IEEE Photon Technology Letter, vol. 14, no. 2, pp. 155, 2002.
3Y. J. Wen, "Power level optimization of 40 Gb/s DWDM systems with hybrid Raman/EDFA amplification", In Proc. Conf. Optical Internet/Australian Conf. Optical Fibre Technology (COIN/ACOFT), Melbourne, Australia, pp. 309, 2003.
4K. I. Kitayama, "Ultimate performance of optical DSB signal-based millimeter- wave fiber-radio system: Effect of laser phase noise", Journal of Lightwave Technology, vol. 17, no. 10, pp. 1774, 1999.
5J. R. Barry, and E. A. Lee, "Performance of coherent optical receivers", Proc Inst Elect Eng, vol. 78, no. 8, pp. 1369, 1990.
6J. Salz, "Modulation and detection for coherent lightwave communications", IEEE Commun Magazine, vol. 24, no. 6, pp. 38, 1986.
7P. B. Gallion, and G. Debarge, "Quantum phase noise and field correlation in single frequency semiconductor laser systems", IEEE J Quantum Electron, vol. 6, no. 4, pp. 343, 1984.
8R. W. Tkach, "Phase noise and line width in an InGaAsP DFB laser", J Lightwave Technology, vol. 4, no. 11, pp. 1711, 1986.
9T. S. Cho, C. Yun, J. I. Song, and K. Kim, "Analysis of CNR penalty of Radio-Over-Fiber systems including the effects of phase noise from laser and RF oscillator", Journal of Lightwave Technology, vol. 23, no. 12, 2005.
10V. Sharma, A. Singh, and A. K. Sharma, "Analysis and simulation of the effect of spectral width over intensity noise under the impact of higher-order dispersion parameters in the optical communication systems", Opt Commun, vol. 28, no. 1, pp. 3495, 2008.
11V. Sharma, A. Singh, and A. K. Sharma, "Simulative investigation on the impact of laser-spectral width in single-tone radio-over-fiber transmission system using optical single side-band technique", Optics and Lasers in Engineering, vol. 47, pp. 1145, 2009.
12U. Gliese, "Chromatic dispersion in fiber-optic microwave and millimeter-wave links", IEEE Trans Microwave Theory Tech, vol. 44, no. 10, pp. 1716, 1996.
13T. S. Cho, "Modified BER estimation of optical link considering the nonlinear effect of the fiber", Korean Inst Commun Sci (KICS), vol. 23, no. 2, pp. 1707, 2001.
14M. Iqbal, and K. Kim, "Performance of QAM systems with respect to phase noise spectral shape", Proceedings of the Canadian conference on electrical and computer engineering, Halifax, NS, Canada, pp. 14-16, 2000.
15M. Iqbal, J. Lee, and K. Kim, "Performance comparison of digital modulation schemes with respect to phase noise spectral shape", Proceedings of the electrical and computer engineering, Halifax, NS, Canada, pp. 85, 2000.
16K. Ghairabeh, K. Gard, and M. Steer, "Accurate Estimation of Digital Communication System Metrics - SNR, EVM in a Nonlinear Amplifier Environment", IEEE Transactions on Communications, pp. 734, 2005.