Expand this Topic clickable element to expand a topic
Skip to content
Optica Publishing Group

Performance improvement of double-sideband signals in radio-over-fiber links utilizing pre-distortion method

Open Access Open Access

Abstract

In this paper, we investigate the optimal carrier-to-sideband ratio (CSR) for optical double-sideband (DSB) signals in radio-over-fiber (RoF) transmission. A pre-distortion method based on spectral shaping is proposed to optimize the signal CSR and eliminate the dispersion-induced power fading. A 12 GHz RoF transmission over 29 km standard single mode fiber (SSMF) is experimentally demonstrated which reveals that pre-distorted DSB has a 4.4 dB improvement over the one without pre-distortion, and 1.2 dB sensitivity advantage over single sideband (SSB) both with optimal CSR. The pre-distortion method is also applied to multi-channel transmission. The multi-channel experiment shows that for all channels the dispersion-induced power fading effects can be simultaneously mitigated and the pre-distorted DSB signals have ~1 dB sensitivity improvement over the SSB signals.

©2014 Optical Society of America

1. Introduction

As wireless access network advances toward higher capacity to deliver a wide variety of broadband services, radio-over-fiber (RoF) technique has attracted much attention due to the inherent large bandwidth and low-loss of optical fiber transmission [15]. One of the key elements that determine the performance of a RoF link is the optical modulation of RF signals, especially in the microwave and millimeter-wave bands [68]. Among all the approaches, optical microwave signal generation employing external modulation is considered as the simplest and most cost-effective method [6]. By directly modulating the wireless signals onto an optical carrier, optical microwave signals with double sidebands (DSB) can be obtained. However, the optical DSB signal is susceptible to the fiber dispersion that induces different phase shifts to the two sidebands and subsequently leads to power fading of the received RF signals [9]. Much work has been carried out to overcome this fiber dispersion effect, where the most straightforward technique aims to remove one spectral component, including the single sideband (SSB) and optical carrier suppression (OCS) modulation schemes [5,6,912]. OCS has an inherent advantage in power balance between two sidebands which leads to a sensitivity improvement. However, this technique requires a large RF drive power to obtain a desirable modulation depth as the modulator works in the nonlinear region [5]. Moreover, due to the bit-walk-off effects, OCS can hardly perform better than SSB for a long transmission distance under the condition of optimal carrier to sideband ratio (CSR).

Another issue related to optical RF signals is the weak modulation depth. Due to the limited linear dynamic range of intensity modulator, the RF signals with high frequency are usually weakly modulated onto the optical carrier. A few techniques have been proposed to improve the modulation efficiency of the optical RF signals [13,14]. Amongst these approaches, employing fiber grating as an external filter to decrease the carrier ratio provides obvious advantages of simplicity and low cost [14]. However, this method lacks flexibility for use in RoF systems since the central wavelength of a fiber grating can be hard to adjust in a broad wavelength range.

In this paper, we use two approaches to improve the transmission performance of optical DSB signals. First, we theoretically show that optimal CSR for DSB signals is 3 dB and this optimal CSR can be achieved via filtering process. Secondly, following the trend in high-speed digital transmission where pre-distortion is employed to remove channel dispersion [15,16], we propose pre-distortion by adding beforehand phase shift between the two sidebands of the transmitted signals to cancel the dispersion effects for ROF systems. Compared to the conventional dispersion compensation methods, such as using chirped fiber gratings or dispersion compensation fiber after transmission [17], the pre-distortion method achieves better flexibility as it can be implemented by an adaptive optical Waveshaper or even high-speed electronic DSP and digital-to-analogue-convertors (DAC) (currently at an analogue bandwidth of 15 GHz and to be doubled to 30 GHz in the next decade) [18]. Moreover, it can be deployed in central office to simplify the configuration of base stations in RoF systems. Furthermore, when using the Waveshaper to process the signal spectra in frequency domain, simultaneous multi-channel manipulation and flexible system reconfiguration can be achieved. The experimental results show that by employing this pre-distortion method based on spectral shaping, the receiver sensitivity at the BER of 10−3 can be improved by 4.4 dB after 29 km transmission for DSB signals with 12 GHz RF frequency. Moreover, the pre-distorted DSB signals with an optimal CSR of 3 dB achieve a 1.2 dB sensitivity improvement over SSB signals with an optimal CSR of 0 dB. Multi-channel transmission has also been demonstrated to verify the effectiveness of the proposed method in multi-channel scenario. Our method can be well extended to higher RF frequency, since the adaptive Waveshaper exhibits better performance under the condition of large frequency spacing which is well above the resolution limitation of the Waveshaper.

2. Concept of pre-distortion for optical DSB signals

Two pre-distortion procedures have been applied to the optical DSB signals to improve the performance for RoF transmission, including CSR optimization and dispersion pre-compensation. Because unavailability of high-speed DAC, we achieve dispersion pre-compensation by adding beforehand phase shift for different spectral components using a Waveshaper, which is essential for performance improvement of optical DSB signals after a fiber transmission. Meanwhile, the Waveshaper can be faultlessly used as a tunable external optical filter to optimize the CSR of optical RF signals, since it can provide a considerable frequency resolution.

Figure 1 shows the conceptual diagram of the proposed pre-distortion method using a Waveshaper, a powerful optical processor made of gratings and solid-state liquid crystal on silicon (LCOS) engines (Finisar 4000S). Firstly, the spectral components of the DSB signals are split out spatially through a grating. Then different attenuation and phase shift are added onto each frequency component to adjust the CSR and add preset phase shift. After being combined by another grating, the signal pre-distortion is achieved. The frequency resolution of this Waveshaper is 10 GHz. Actually line-by-line spectral shaping techniques have achieved 357-MHz resolution [19]. Besides the function of phase shift and attenuation, it also can be extended to manipulate multi-channels, which makes Waveshper suitable for the configuration in central offices. The concepts and derivation of the optimal CSR and dispersion pre-compensation are presented as below.

 figure: Fig. 1

Fig. 1 Conceptual diagram of pre-distortion method for optical DSB signals using a Waveshaper.

Download Full Size | PDF

2.1 CSR optimization

The optimal CSR for optical SSB signals has been investigated to be 0 dB [14]. This value for optical DSB signals can be derived to be 3 dB by using an analogous method. Define the received optical DSB signal power to be Prec and the ratio of each sideband power to the total power to be α, then the optical power of each sideband and the optical carrier are αPrec and (12α)Prec respectively. According to the principle of direct detection, the received signal current can be described by

Isig2α(12α)Prec,0α1.
It is easy to find the maximum received signal current is attained when αequals 0.25, which means the optimal CSR for optical DSB signals is 3 dB. It can be also shown that both at optimal CSR condition, the receiver sensitivity of DSB is 1.5 and 3.0 dB better than SSB for direct and pre-amplified detection, respectively.

2.2 Dispersion pre-compensation

In RoF systems using external optical modulation technique, the output signal of the modulator can be expressed as [20]:

E(t)cos[UbVππ2+URFVππ2cos(2πfRFt)]cos(2πf0t)
where Ub is the bias voltage of the modulator. URF and fRF denote the amplitude and the frequency of the RF signals, respectively. f0 is the frequency of the optical carrier. The optical field after the dispersion pre-compensation and the fiber link can be derived by using Bessel functions as
E(t)J0(m)cos(2πf0t+φ0+φ0)J1(m){cos[2π(f0fRF)t+φ1+φ1]+cos[2π(f0+fRF)t+φ2+φ2]}+...
where φ0, φ1 and φ2 represent the different phase delays of the spectral components at the carrier, lower and upper sidebands due to the chromatic dispersion of the fiber link. φ0, φ1and φ2 are the introduced phase shift by the Waveshaper. m=URF/Vπ is the normalized amplitude of the RF signals.

At the photodiode end, the receiving power at the frequency fRF is

Pfcos2[φ1+φ1+φ2+φ22(φ0+φ0)]=cos2[(φ1+φ22φ0)2+(φ1+φ22φ0)2].
Actually the phase shift at each frequency is a relative value which changes with the reference frequency. Here we set the carrier frequency as the reference for simplicity. Then we have φ0=φ0=0, φ1=φ2 and φ1=φ2. Equation (4) can be expressed as [5]
Pf=cos2(πDLλ2cfRF2+φ1)
where D and L denotes the dispersion and length of the transmission fiber, λ is the wavelength of the optical carrier and c is the vacuum light velocity. It can be seen from Eq. (5) that in order to cancel the power fading effect, we need to set

φ0=0andφ1=φ2=πDLλ2cfRF2

In order to analyze the influence of the phase shift induced by the Waveshaper, we define the phase shift unit as

Pu=πDLλ2c
where the accuracy of D and L affects the performance of pre-distorted DSB signals.

3. Experiment and results

As mentioned above, the Waveshaper is suitable to be deployed in central office to manipulate multi-channels simultaneously, including generating SSB signals by removing one sideband, controlling the signal CSR and compensating fiber dispersion by pre-distortion. In this section, the effectiveness of the proposed pre-distortion method for single channel is firstly investigated. Then the multi-channel transmission using the pre-distorted DSB signals is demonstrated.

3.1 Single channel transmission

The experimental setup for optical pre-distorted DSB and SSB signals generation and transmission is shown in Fig. 2. Firstly 1 Gb/s binary on-off keyed data streams (215-1 pseudorandom binary sequence) are up-converted to 12 GHz microwave frequency and modulated onto an optical carrier with the power of 13 dBm at a wavelength of 1549 nm to generate the DSB signals as shown in inset (i). The optical modulator is biased at linear mode and the initial CSR of the DSB signals is 22dB. The original DSB signals are then passed through a Waveshaper to generate optical SSB signals or pre-distorted DSB signals with desired CSRs, as shown in insets (ii) and (iii) respectively. After the Waveshaper, signals are amplified using an erbium-doped fiber amplifier (EDFA) and sent into a 29 km fiber. At the receiver end the signals are detected with a high speed real-time oscilloscope for offline processing, where the DSP procedures including down converting the signals to baseband followed by a LPF. We firstly investigate the optimal values of the CSR for optical DSB and SSB signals. Figure 3 depicts the sensitivity at BER = 10−3 for a range of CSR for both signals. It can be seen that the optimal CSR for optical DSB and SSB signals are 3 and 0 dB, respectively, which shows a good agreement with the theoretical analysis.

 figure: Fig. 2

Fig. 2 Experimental setup for a pre-distorted ROF system. (IM: intensity modulator; WS: Waveshaper; PD: photodiode; LPF: low pass filter). Insets (i), (ii) and (iii) are the initial DSB signal spectrum, pre-distorted DSB and SSB signal spectra with optimal CSR, respectively.

Download Full Size | PDF

 figure: Fig. 3

Fig. 3 Required received optical power for a BER of 10-3 versus CSR for DSB and SSB signals.

Download Full Size | PDF

Figure 4(a) shows the BER performance for optical DSB and SSB signals with different CSRs. Compared with the SSB signals with 0 dB CSR, the sensitivity of DSB signals with optimal CSR has a 1.5 dB improvement for the back to back transmission as expected. Even with a non-optimal CSR (0, 6 dB), the DSB signals exhibit a sensitivity advantage over SSB signals with optimal CSR under the back-to-back scenario. To verify the effectiveness of the phase control function of the Waveshaper and ensure the BER curve for optical DSB signals are measured without dispersion effects, we artificially add phase shift onto the back-to-back DSB signals. Figure 4(b) shows the received power at BER of 10−3 as the function of the added phase shift unit Pu.

 figure: Fig. 4

Fig. 4 (a) Back-to-back BER performance for optical SSB and DSB signals with different CSRs; (b) Required received power at the BER of 10−3 versus unit phase shift per GHz2 for back-to-back DSB signals with 3 dB CSR.

Download Full Size | PDF

We next demonstrate the effectiveness of the pre-distortion method for optical DSB signals after 29 km transmission. As shown in Fig. 5(a), the red curve indicates the BER of SSB signals which shows that the sensitivity is nearly the same as the signals under back to back transmission as it is immune to the dispersion-induced power fading effects. The blue curve shows BER of the conventional DSB signals with 3 dB CSR, which suffer from severe performance degradation due to dispersion induced power fading. It can be seen that the pre-distorted DSB signals (green curve) achieve a 4.4 dB improvement over the one without pre-distortion. Moreover, it also has a 1.2 dB sensitivity advantage over the optical SSB signals with optimal CSR. Since the inaccuracy of the fiber dispersion, length and the Waveshaper precision can affect the performance of the pre-distorted DSB signals, we have varied the phase shift unit of the Waveshaper near the theoretical value (0.0124 rad/GHz2) to achieve an optimal performance. Figure 5(b) illustrates how the sensitivity varies with the added phase shift. It confirms the fact that an over-distortion will result in a power fading and degraded the transmission performance.

 figure: Fig. 5

Fig. 5 (a) BER curve for SSB signals with 0 dB SCR, DSB signals with 3 dB CSR with and without pre-distortion after 29 km transmission; (b) Required received power at the BER of 10−3 versus unit phase shift per GHz2 for the Waveshaper setting.

Download Full Size | PDF

3.2 multi-channel transmission

Besides the flexible spectral shaping for the single channel, another obvious advantage of using the Waveshaper is the simultaneous manipulation of multi-channels, which can bring a significant reduction in the cost for each channel. We investigate the feasibility of the proposed approach for multi-channel transmission. Figure 6 shows the experimental setup. Three laser diodes (LD) with different wavelengths (1549.5, 1549.9, 1550.3 nm) are used to generate three channels with 50GHz frequency spacing. Then they are modulated by the same microwave signals with 1Gbit/s data rate and 12 GHz frequency. The Waveshaper is employed subsequently to implement pre-distortion scheme onto multi-channel signals, including SSB signal generation, CSR optimization and dispersion pre-compensation. The insets (i) and (ii) of Fig. 6 respectively show the generated DSB and SSB signals with optimal CSR (Three channels are labeled as C1, C2 and C3 respectively). A 29 km SMF is used to transmit the multi-channel signals. The receiving end of the multi-channel transmission is the same as the configuration of single channel system. We also use off-line DSP method to investigate the system performance.

 figure: Fig. 6

Fig. 6 Experimental setup of multichannel transmission. Insets (i) and (ii) are the DSB and SSB signal spectra with optimal CSR respectively.

Download Full Size | PDF

Then we measure the performance of the RoF signals with the wavelength of 1549.9 nm, which is the middle channel (C2). As shown in Fig. 7(a), the red curve is the BER of SSB signals over 30 km transmission, which shows that the performance keeps nearly the same level as the signal channel. The blue curve indicates the BER of DSB signals without pre-distortion, which severely suffers from the dispersion-induced power fading effects. The green curve shows the BER of pre-distorted DSB signals. It can be seen that the predistortion method based on spectral shaping is effective for multi-channel transmission. 4.2 dB improvement for the sensitivity of DSB signals has been achieved. Figure 7(b) shows the sensitivities of three channels at BER = 10−3. The performance degradations of three channels caused by dispersion-induced power fading effects are obviously mitigated simultaneously. It can be seen that by pre-distortion processing based on spectral shaping, DSB signals have ~1 dB sensitivity advantages over the SSB signals for all channels.

 figure: Fig. 7

Fig. 7 (a) BER curve of the middle channel for SSB signals with 0 dB CSR, DSB signals with 3 dB CSR with and without pre-distortion in multi- channel transmission; (b) Required received power at the BER of 10−3 for each channel.

Download Full Size | PDF

4. Conclusion

Fading-resilient transmission of 12 GHz optical pre-distorted DSB signals over a 29 km fiber transmission has been experimentally demonstrated. It is shown that the optimal CSR for optical DSB signals is 3 dB. The proposed pre-distortion method achieves a 4.4 dB improvement for optical DSB signals that suffer from fiber dispersion effects and the pre-distorted DSB signals have a 1.2 dB sensitivity improvement over SSB signals both under the optimal CSR condition. Moreover, the proposed pre-distortion method based on spectral shaping is proven effective in multi-channel transmission.

Acknowledgments

The research is supported by the National Basic Research Program of China (2012CB315704), the Natural Science Foundation of China (No. 61275068, 61325023, 61335005) and the Key Grant Project of Chinese Ministry of Education (No.313049).

References and links

1. J. Seeds and K. J. Williams, “Microwave photonics,” J. Lightwave Technol. 24(12), 4628–4641 (2006). [CrossRef]  

2. J. Capmany and D. Novak, “Microwave photonics combines two worlds,” Nat. Photonics 1(6), 319–330 (2007). [CrossRef]  

3. J. P. Yao, “Microwave photonics,” J. Lightwave Technol. 27(3), 314–335 (2009). [CrossRef]  

4. M. Sauer, A. Kobyakov, and J. George, “Radio over fiber for picocellular network architectures,” J. Lightwave Technol. 25(11), 3301–3320 (2007). [CrossRef]  

5. C. Lim, T. A. Nirmalathas, M. Bakaul, P. Gamage, K.-L. Lee, Y. Yang, and R. Waterhouse, “Fiber-wireless networks and subsystem technologies,” J. Lightwave Technol. 28(4), 390–405 (2010). [CrossRef]  

6. J. Yu, Z. Jia, L. Yi, Y. Su, G.-K. Chang, and T. Wang, “Optical millimeter-wave generation or up-conversion using external modulators,” IEEE Photon. Technol. Lett. 18(1), 265–267 (2006). [CrossRef]  

7. D. Fonseca, A. V. T. Cartaxo, and P. Monteiro, “Optical single-sideband transmitter for various electrical signaling formats,” J. Lightwave Technol. 24(5), 2059–2069 (2006). [CrossRef]  

8. Z. Xu, X. Zhang, and J. Yu, “Frequency upconversion of multiple RF signals using optical carrier suppression for radio over fiber downlinks,” Opt. Express 15(25), 16737–16747 (2007). [CrossRef]   [PubMed]  

9. U. Gliese, S. Norskov, and T. N. Nielsen, “Chromatic dispersion in fiber-optic microwave and millimeter-wave links,” IEEE Trans. Microw. Theory Tech. 44(10), 1716–1724 (1996). [CrossRef]  

10. Y. Cui, K. Xu, J. Dai, X. Q. Sun, Y. T. Dai, Y. F. Ji, and J. T. Lin, “Overcoming chromatic-dispersion-induced power fading in ROF links employing parallel modulators,” IEEE Photon. Technol. Lett. 24(14), 1173–1175 (2012). [CrossRef]  

11. S. Li, X. Zheng, H. Zhang, and B. Zhou, “Compensation of dispersion-induced power fading for highly linear radio-over-fiber link using carrier phase-shifted double sideband modulation,” Opt. Lett. 36(4), 546–548 (2011). [CrossRef]   [PubMed]  

12. G. H. Nguyen, J. Poette, and B. Cabon, “Importance of Chirp Effect in Millimeter Wave Optical Upconversion Systems,” J. Lightwave Technol. 29(12), 1753–1758 (2011). [CrossRef]  

13. S. Tonda-Goldstein, D. Dolfi, J.-P. Huignard, G. Charlet, and J. Chazelas, “Stimulated brillouin scattering for microwave signal modulation depth increase in optical links,” Electron. Lett. 36(11), 944–946 (2000). [CrossRef]  

14. C. Lim, M. Attygalle, A. Nirmalathas, D. Novak, and R. Waterhouse, “Analysis of Optical Carrier-to-Sideband Ratio for Improving Transmission Performance in Fiber-Radio Links,” IEEE Trans. Microw. Theory Tech. 54(5), 2181–2187 (2006). [CrossRef]  

15. R. I. Killey, P. M. Watts, V. Mikhailov, M. Glick, and P. Bayval, “Electronic dispersion compensation by signal predistortion using digital processing and a dual-drive Mach-Zehnder modulator,” IEEE Photon. Technol. Lett. 17(3), 714–716 (2005). [CrossRef]  

16. D. McGhan, C. Laperle, A. Savchenkov, C. D. Li, G. Mak, and M. O'Sullivan, “5120 km RZ-DPSK transmission over G652 fiber at 10 Gb/s without optical dispersion compensation,” IEEE Photon. Technol. Lett. 18(2), 400–402 (2006). [CrossRef]  

17. T. A. Birks, D. Mogilevtsev, J. C. Knight, and P. St. J. Russell, “Dispersion compensation using single material fibers,” IEEE Photon. Technol. Lett. 11, 674–676 (1999). [CrossRef]  

18. P. W. Juodawlkis, J. C. Twichell, G. E. Betts, J. J. Hargreaves, R. D. Younger, J. L. Wasserman, F. J. O’Donnell, K. G. Ray, and R. C. Williamson, “Optically sampled analog-to-digital converters,” IEEE Trans. Microw. Theory Tech. 49(10), 1840–1853 (2001). [CrossRef]  

19. J. T. Willits, A. M. Weiner, and S. T. Cundiff, “Line-by-line pulse shaping with spectral resolution below 890 MHz,” Opt. Express 20(3), 3110–3117 (2012). [CrossRef]   [PubMed]  

20. H. Schmuck, “Comparison of optical millimeter-wave system concepts with regard to chromatic dispersion,” Electron. Lett. 31(21), 1848–1849 (1995). [CrossRef]  

Cited By

Optica participates in Crossref's Cited-By Linking service. Citing articles from Optica Publishing Group journals and other participating publishers are listed here.

Alert me when this article is cited.


Figures (7)

Fig. 1
Fig. 1 Conceptual diagram of pre-distortion method for optical DSB signals using a Waveshaper.
Fig. 2
Fig. 2 Experimental setup for a pre-distorted ROF system. (IM: intensity modulator; WS: Waveshaper; PD: photodiode; LPF: low pass filter). Insets (i), (ii) and (iii) are the initial DSB signal spectrum, pre-distorted DSB and SSB signal spectra with optimal CSR, respectively.
Fig. 3
Fig. 3 Required received optical power for a BER of 10-3 versus CSR for DSB and SSB signals.
Fig. 4
Fig. 4 (a) Back-to-back BER performance for optical SSB and DSB signals with different CSRs; (b) Required received power at the BER of 10−3 versus unit phase shift per GHz2 for back-to-back DSB signals with 3 dB CSR.
Fig. 5
Fig. 5 (a) BER curve for SSB signals with 0 dB SCR, DSB signals with 3 dB CSR with and without pre-distortion after 29 km transmission; (b) Required received power at the BER of 10−3 versus unit phase shift per GHz2 for the Waveshaper setting.
Fig. 6
Fig. 6 Experimental setup of multichannel transmission. Insets (i) and (ii) are the DSB and SSB signal spectra with optimal CSR respectively.
Fig. 7
Fig. 7 (a) BER curve of the middle channel for SSB signals with 0 dB CSR, DSB signals with 3 dB CSR with and without pre-distortion in multi- channel transmission; (b) Required received power at the BER of 10−3 for each channel.

Equations (7)

Equations on this page are rendered with MathJax. Learn more.

I sig 2 α(12α) P rec , 0α1.
E(t)cos[ U b V π π 2 + U RF V π π 2 cos(2π f RF t)]cos(2π f 0 t)
E(t) J 0 (m)cos(2π f 0 t+ φ 0 + φ 0 ) J 1 (m){cos[2π( f 0 f RF )t+ φ 1 + φ 1 ]+cos[2π( f 0 + f RF )t+ φ 2 + φ 2 ]}+...
P f cos 2 [ φ 1 + φ 1 + φ 2 + φ 2 2 ( φ 0 + φ 0 )]= cos 2 [ ( φ 1 + φ 2 2 φ 0 ) 2 + ( φ 1 + φ 2 2 φ 0 ) 2 ].
P f = cos 2 ( πDL λ 2 c f RF 2 + φ 1 )
φ 0 =0 and φ 1 = φ 2 = πDL λ 2 c f RF 2
Pu= πDL λ 2 c
Select as filters


Select Topics Cancel
© Copyright 2024 | Optica Publishing Group. All rights reserved, including rights for text and data mining and training of artificial technologies or similar technologies.