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

Cost-effective colorless WDM-PON delivering up/down-stream data and broadcast services on a single wavelength using mutually injected Fabry-Perot laser diodes

Open Access Open Access

Abstract

We propose and demonstrate for the first time a cost-effective and colorless wavelength-division-multiplexed passive optical network which can provide a full-duplex 155-Mb/s up/downstream data and broadcast signals on a single wavelength using light-injected optical sources and bidirectional transmission in a single fiber. The optical sources for downstream and upstream baseband data are Fabry-Perot laser diode (FPLD) at the central office and a reflective semiconductor optical amplifier (RSOA) at each optical network unit, respectively. The broadband light source (BLS) for the downstream is implemented by using mutually injected FPLDs. The transmission experiment shows that the error-free operation for baseband data and a high-quality broadcast signal with 3-dB CNR margin can be obtained.

©2008 Optical Society of America

1. Introduction

Wavelength-division-multiplexed passive optical networks (WDM-PONs) have been considered as one of the strong candidates for the future access networks thanks to their large capacity, strong security, and high flexibility. However, their initial deployment and maintenance are highly expensive due to a number of wavelength-specified light sources. In addition, the difficulty in providing broadcasting service caused by logical point-to-point connection between central office (CO) and optical network units (ONUs) is one of the reasons that retard the progress on practical utilization of WDM-PON. To overcome this problem on the one hand, many techniques have been proposed to develop a low-cost, colorless light sources including Fabry-Perot laser diodes (FPLDs) and reflective semiconductor optical amplifiers (RSOAs), both externally injected by incoherent broadband light source (BLS) for colorless operation [1]–[4]. However, the BLS, implemented in general by either a super luminescent diode or amplified spontaneous emission (ASE) in general, is still bulky and expensive, and its high relative intensity noise (RIN) limits the maximum data rate of the signals [5]. Recently, a coherent BLS has been proposed using mutually injected FPLDs (MI-FPLD) to reduce the cost and RIN of BLS [6].

On the other hand, there has recently been a strong demand for so-called “triple-play,” broadband access services including the Internet, Internet Protocol (IP) telephony, and video broadcasting. A video overlay technique has been also proposed in a WDM-PON system using an MI-FPLD as a BLS for broadcasting and upstream transmission [7]. Unlike other video overlay techniques, this does not require the modification of either remote node (RN) at outdoor or the transmitters at CO to deliver the broadcast signals to ONU. However, this technique should employ an external modulator devoted to the broadcast signals. In addition, since this technique assigned a dedicated wavelength band to the broadcast signals, the ONU should be able to accommodate three wavelengths for the broadcasting and up/downstream baseband data services in the worst case. These, ultimately, increase the system cost and make worse bandwidth utilization, which weaken the original purpose of developing a low-cost and colorless light source for WDM-PONs.

In this paper, we propose and demonstrate for the first time a cost-effective and colorless WDM-PON which can deliver up/downstream baseband data as well as broadcasting services on a single wavelength. We have already demonstrated the transmission of broadcast and upstream signals using an MI-FPLD in [8]. To improve the wavelength utilization of the network, we adopt the bidirectional transmission in a single fiber for up/downstream baseband data and video services. This not only reduces the required number of light sources and wavelengths but also relieves the complexity of the transceivers in the WDM-PON by removing or reducing the number of optical filters for the separation of the multiple services. For better the cost-effectiveness, the proposed network employs 1) a cost-effective BLS based on MI-FPLD, 2) a bidirectional transmission technique to minimize the fiber count, and 3) low-cost FPLDs and RSOAs for modulation of downstream and upstream baseband data, respectively. In addition, we directly modulate one of the FPLDs in MI-FPLD for video signal transmission instead of using an external modulator, which further reduces the system cost. The MI-FPLD at the CO and a gain-saturated RSOA at each optical network unit (ONU) allow us to utilize loop-back configuration for bidirectional transmission, since it offers low RIN as well as high robustness to Rayleigh back-scattering. Performance of the proposed network is evaluated by transmitting the 155-Mb/s symmetric baseband data in up/downstream and two downstream broadcast signals operating at 1.0 and 1.006 GHz over 20-km standard single mode fiber (SSMF) link. The experimental results show that the error free operation for baseband data and a high-quality broadcast signal with 3-dB CNR margin can be obtained. Since the multi-channel transmission of the WDM-PON utilizing the MI-FPLD and FPLD for the BLS and data modulation, respectively, has been already demonstrated in [6], we present a single wavelength transmission in this paper for the sake of simplicity.

2. Operational Principle

 figure: Fig. 1.

Fig. 1. The schematic diagram of the proposed WDM-PON.

Download Full Size | PDF

Figure 1 shows the principle of the proposed WDM-PON. The MI-FPLD, which consists of two FPLDs (FPLDa and FPLDb), a coupler, two isolators, and a polarization beam combiner (PBC) is used as a BLS for the downstream transmission. The FPLDs are connected with each other using the coupler for mutual injection and the outputs of the coupler were polarization-multiplexed to make the MI-FPLD output unpolarized. The output of the MI-FPLD, which produces multiple longitudinal modes, is fed to a circulator and then mode-sliced by an arrayed-waveguide grating (AWG) used for multiplexing the downstream signals at the CO. Since the mode-spacing of the MI-FPLD is the same as the channel spacing of the AWG, only one spectrally-sliced single-mode of MI-FPLD can be injected to each FPLDi, where i=1 to N. Obviously, the wavelength of each downstream signal is determined by that of injected mode from MI-FPLD. The broadcast signals are applied to the MI-FPLD by directly modulating of one of the two FPLDs in MI-FPLD. Thus, the broadcast signals can be transmitted to every ONU along with the downstream signals simultaneously. The modulated downstream signals are wavelength-multiplexed by the AWG, transmitted through a circulator and feeder fiber, wavelength-demultiplexed by an AWG in the RN, and sent to the corresponding ONU.

In ONU, a half of the input signal is detected by a photo-detector (PD) for the reception of the downstream and broadcast signals. To separate the baseband data and broadcast signals, the electrical signal is split two ways and then filtered with an electrical low-pass filter (LPF) for the downstream data signal and a band-pass filter (BPF) for the broadcast signal. The other half of optical input signal is injected to the RSOA for the remodulation of RSOA with the upstream baseband data. The RSOA operating at the gain-saturation region can squeeze out the downstream baseband data, and enables the upstream data to be imposed upon the downstream signal directly [9]. The remodulated upstream signal is transmitted back to the CO via the RN, feeder fiber, and circulator, and then demultiplexed by an AWG. At the CO, the unsuppressed broadcast signals are removed from the upstream baseband data by an LPF of the receiver.

3. Experiment

Figure 2 shows the experimental setup. The center wavelength and mode spacing of FPLDs used for both MI-FPLD and downstream modulation are 1545 nm and 0.8 nm, respectively. Since the mode spacing of these FPLDs is 0.8 nm, the output modes of the MI-FPLD can be well matched to the frequency spacing of the ITU grid. In addition, the center wavelength of the envelope can be accurately adjusted by temperature control of the FPLDs. The reflectivity of front facet of them was intentionally reduced to be less than 1 % by using anti-reflection coating for the efficient injection of external light. The output power and center wavelength of MI-FPLD at the bias current of 40 mA were +6 dBm and 1552.028 nm after mutual injection respectively. Two tone signals representing the broadcast signals at 1.0 and 1.006 GHz were combined with an RF combiner and directly applied to FPLDa. To make the polarizations of two FPLDs orthogonal state in the MI-FPLD, we used the polarization maintaining fiber (PMF) between the coupler and PBC. The AWGs located at CO in Fig. 1 were replaced with a cascade of 0.6-nm-bandwidth optical bandpass filters. The injected optical power to the FPLD used for downstream modulation was -15.6 dBm after mode slicing and it was modulated with 155-Mb/s downstream baseband data. The spectral bandwidth of downstream baseband data was limited by a low-pass filter with 466-MHz bandwidth before applying to FPLD1 to avoid the interference between the downstream data and broadcast signals. The bias current of FPLD1 and optical power of the modulated downstream signal measured at EDFA input were 38 mA and -10 dBm, respectively. The modulated downstream signal was amplified by the EDFA to +3 dBm/channel, transmitted over 20-km SSMF, demultiplexed by an optical bandpass filter with 1-nm bandwidth in the RN, and sent to the ONU. Even though we used an EDFA for the deep gain-saturation of the RSOA, it can be removed in commercial systems to improve the cost-effectiveness after optimizing the gain-saturation characteristics of the RSOA and coupling ratio of the optical coupler in the ONU. The half of the optical signal to ONU was divided by a 3-dB optical coupler and then one of outputs was detected by a PD to receive the downstream data and broadcast signals. The performance of each signal was evaluated by measuring the BER and CNR with an error detector (ED) and an RF spectrum analyzer (RFSA) for downstream baseband data and broadcast signals, respectively. The other half was injected to the RSOA and remodulated by 155-Mb/s upstream baseband data. We placed a polarization controller at the input of RSOA, but it can be removed by using a polarization-insensitive RSOA. The remodulated upstream signal was detected at the CO. In our experiment, the 3-dB bandwidth of the baseband receivers for downstream and upstream were 117 MHz which would make the baseband signal free from the effect of un-suppressed broadcast signals. Finally, the BER performance of the baseband signal was measured with the ED.

 figure: Fig. 2.

Fig. 2. Experimental setup. OBPF: optical bandpass filter, PC: polarization controller

Download Full Size | PDF

4. Results and Discussions

Figure 3(a) shows the optical spectra of the two FPLDs used in the MI-FPLD (i.e. FPLDa and FPLDb) before mutual injection. The frequency difference between two FPLDs was 0.14 nm, and the output power of each FPLD was -3 dBm at a bias current of 40 mA. After mutual injection, the output power of BLS was increased to +6 dBm and we could obtain 11 modes within 10-dB bandwidth as shown in Fig. 3(b). In addition, the center wavelength of the envelope was shifted to long wavelength due to the reduced threshold current of the MI-FPLD after mutual injection [6]. The 3-dB linewidth of each mode was increased from 0.08 nm to 0.14 nm after mutual injection. The side-mode suppression ratio, after mode-slicing, was measured to be slightly less than 20 dB due to the shallow roll-off after 3-dB cut-off frequency of the OBPFs. However, it was improved to 28 dB after injection to the FPLD1 as shown in Fig. 3(c), which is large enough to ignore the effects of fiber dispersion in WDM-PON applications. Figure 3(c) also shows that the side modes of downstream signal are significantly suppressed and veiled with amplified spontaneous emission (ASE) noise of EDFA and RSOA after remodulation at the ONU.

 figure: Fig. 3.

Fig. 3. Measured optical spectra of (a) FPLDs used in MI-FPLD before mutual injection, (b) output of MI-FPLD without and with mode-slicing after mutual injection, and (c) downstream and upstream signals measured at EDFA and PD input in the CO, respectively.

Download Full Size | PDF

 figure: Fig. 4.

Fig. 4. RIN of MI-FPLD in comparison with that of commercial tunable laser diode and a free-running FPLD.

Download Full Size | PDF

Figure 4 shows the measured RIN of MI-FPLD in comparison with that of a commercial tunable laser diode (TLD) and a free-running FPLD. The RIN of the free-running FPLD without mode-slicing was measured to be >-120 dB/Hz throughout the measured frequency range. However, when we performed mutual injection of FPLDs, the RIN was considerably reduced to as low as -150 dB/Hz which is similar to the RIN of the TLD. In this case, however, we observed strong RIN peaks periodically spaced at ~300 MHz. This was because, for the efficient injection of external light, we intentionally reduced the reflectivity of the front facet in each FPLD (<1%), which facilitates the formation of a new cavity between two rear facets of FPLDs. In our MI-FPLD, the fiber length between FPLDa and FPLDb is about 33 cm, which corresponds to a free-spectral range of 300 MHz. Nevertheless, we achieved the RIN of <-140 dB/Hz even after mode-slicing at wide frequency ranges between the peaks. Obviously, this rise in RIN should be attributed to the mode partition noise. To prevent the noise peaks from falling in the baseband spectrum, we set the data rate of upstream and downstream signals to be 155 Mb/s.

 figure: Fig. 5.

Fig. 5. Effect of amplitude of broadcast signal on downstream signal (a) broadcast signalinduced power penalty (b) RIN as a function of RF power of broadcast signal.

Download Full Size | PDF

 figure: Fig. 6.

Fig. 6. BER curves of (a) downstream data and (b) upstream data while varying the extinction ratio of downstream signal

Download Full Size | PDF

Since up/downstream data together with the broadcast signals should be delivered on a single wavelength, we need to optimize the modulation depth of each signal. Firstly, to determine the amplitude of the broadcast signals, we measured the power penalty of the downstream signal as we increased the RF power of the broadcast signals. Figure 5(a) shows the measured power penalty of the downstream baseband data as a function of the RF power of the broadcast signals applied to FPLDa. It clearly exhibits that when the RF power increases more than 3 dBm per channel, the penalty rapidly increases. This is because the large RF power of broadcast signals which were directly fed to one of the FPLDs disturbs the stable injection conditions of the MI-FPLD, which, in turn, increases the RIN of the MI-FPLD. Fig. 5(b) shows that when we set the RF power of broadcast signal to be more than 3 dBm per channel, RIN begins to increase sharply. The large RIN ripples at 100, 200, 400, 500, and 700 MHz were generated by the intermodulation between the noise peaks spaced at 300 MHz and the broadcast signals located at ~ 1 GHz and 1.006 GHz. We also observed a power variation in optical spectrum and the number of modes the MI-FPLD can provide within 10-dB bandwidth was reduce to 9 when the RF power was larger than 3 dBm per channel. Thus, we set the RF power of broadcast signals to be 1 dBm per channel to maintain the power penalty less than 0.3 dB. To accommodate more broadcast channels, the RF power of each channel should be reduced to the level that the mutual injection in MI-FPLD is kept stable or we should employ an external modulator at the output of the MI-FPLD for the broadcast signals. Next, we optimized the extinction ratio of the downstream data. In our symmetric full-duplex transmission, the gain-saturated RSOA erases the downstream data and imposes the upstream data onto the amplitude-squeezed downstream light. Therefore, the high extinction ratio (ε) of the downstream data adversely affects the upstream performance since, after the RSOA, it will make the downstream data to remain and interfere with the upstream data.

Figure 6 shows the measured BER of downstream and upstream signals as we vary ε. Obviously, the downstream performance increases with ε, whereas the upstream performance decreases with ε. From these results, we compromised the extinction ratio of the downstream signal to be 6 dB which would cause the power penalty of 2 and 0.8 dB on the downstream and upstream signal, respectively. Figure 7(a) shows the measured BER of the up/downstream signals before and after transmission over 20-km SSMF. Since no noticeable penalties are measured at both the downstream and upstream signals after transmission, the effects of chromatic dispersion and nonlinear effects of transmission fiber can be neglected in our network. The deleterious effects of Rayleigh back-scattering are found to be negligible thanks to the wide linewidth of MI-FPLD and relatively high power of the upstream signal [6]. The CNR of the broadcast signals were measured to be 9 dB within the noise bandwidth of 4 MHz as shown in Fig. 7(b). Since the minimum required CNR for good image quality is 6 dB for an MPEG/QPSK video signal with forward error correction [10]–[11], there is still 3-dB margin to carry the digital video signals.

 figure: Fig. 7.

Fig. 7. Measured (a) BER and (b) electrical spectrum after transmission over 20-km SSMF

Download Full Size | PDF

6. Conclusion

We have proposed and successfully demonstrated a cost-effective WDM-PON which can provide up/downstream data as well as broadcast services on a single wavelength by using a low-cost MI-FPLD as an external injection light source, and a FPLD and an RSOA for downstream and upstream modulators, respectively. We have found that the periodic RIN peaks of the MI-FPLD, which are created by a long cavity between the FPLDs, limit the maximum data rate of the baseband signals. For higher bit rate transmission of baseband signals, the physical length of the cavity should be reduced, thereby achieving wider noise-peak-free region in the spectrum. We also found that the amplitude of broadcast signals affects the RIN of MI-FPLD and it plays a key role in the performance of proposed network as well as extinction ration of downstream signal.

Acknowledgments

This work was supported in part by the Information and Telecommunication National Scholarship Program of the Ministry of Information and Communication (MIC), Korea.

References and links

1. D. J. Shin, Y. C. Keh, J. W. Kwon, E. H. Lee, J. K. Lee, M. K. Park, J. W. Park, Y. K. Oh, S. W. Kim, I. K. Yun, H. C. Shin, D. Heo, J. S. Lee, H. S. Shin, H. S. Kim, S. B. Park, D. K. Jung, S. T. Hwang, Y. J. Oh, D. H. Jang, and C. S. Shim, “Low-cost WDM-PON with colorless bidirectional transceivers,” J. Lightwave Technol. 24, 158–165 (2006). [CrossRef]  

2. P. Healey, P. Townsend, C. Ford, L. Johnston, P. Townley, I. Lealman, L. Rivers, S. Perrin, and R. Moore, “Spectral slicing WDM-PON using wavelength-seeded reflective SOAs,” Electron. Lett. 37, 1181–1182, (2001). [CrossRef]  

3. S. Y. Kim, S. B. Jun, Y. Takushima, E. S. Son, and Y. C. Chung, “Enhanced performance of RSOA-based WDM PON by using Manchester coding,” J. Opt. Netw. 6, 624–630 (2007) http://www.opticsinfobase.org/abstract.cfm?URI=JON-6-6-624 [CrossRef]  

4. H. Kim, S. Kim, S. Hwang, and Y. Oh, “Impact of dispersion, PMD, and PDL on the performance of spectrum-sliced incoherent light sources using gain-saturated semiconductor optical amplifiers,” J. Lightwave Technol. 24, 775–785 (2006). [CrossRef]  

5. J. S. Lee, Y. C. Chung, and D. J. DiGiovanni, “Spectrum-sliced fiber amplifier light source for multichannel WDM applications,” IEEE Photon. Technol. Lett. 5, 1458–1461 (1993). [CrossRef]  

6. K. M. Choi, J. S. Baik, and C. H. Lee, “Broad-band light source using mutually injected Fabry-Perot laser diodes for WDM-PON,” IEEE Photon. Technol. Lett. 17, 2529–2531, (2005). [CrossRef]  

7. J. -H. Moon, K. M. Choi, and C. H. Lee, “Overlay of broadcasting signal in a WDM-PON,” in Proceedings of the Optical Fiber Communication and Conference, (2006), Paper OThK8.

8. H. C. Ji, I. Yamashita, and K. -I. Kitayama, “Transmission of upstream data and broadcast signals over a single wavelength in WDM-PON using light-injected optical sources,” in Proceedings of the European Conference on Optics and Communications, (2007), Paper We6.4.6.

9. W. R. Lee, M. Y. Park, S. H. Cho, J. Lee, C. Kim, G. Jeong, and B. W. Kim, “Bidirectional WDM-PON based on gain-saturated reflective semiconductor optical amplifiers,” IEEE Photon. Technol. Lett. 17, 2460–2462 (2005). [CrossRef]  

10. N. Chand, P. D. Magill, S. V. Swaminathan, and T. H. Daugherty, “Delivery of digital video and other multimedia services (>1 Gb/s bandwidth) in passband above the 155 Mb/s baseband services on a FTTx Full Service Access Network,” J. Lightwve Technol. 17, 2449–2460 (1999). [CrossRef]  

11. D. K. Jung, H. Kim, K. H. Han, and Y. C. Chung, “Spectrum-sliced bidirectional passive optical network for simultaneous transmission of WDM and digital broadcast video signals,” Electron. Lett. 37, 308–309 (2001). [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. The schematic diagram of the proposed WDM-PON.
Fig. 2.
Fig. 2. Experimental setup. OBPF: optical bandpass filter, PC: polarization controller
Fig. 3.
Fig. 3. Measured optical spectra of (a) FPLDs used in MI-FPLD before mutual injection, (b) output of MI-FPLD without and with mode-slicing after mutual injection, and (c) downstream and upstream signals measured at EDFA and PD input in the CO, respectively.
Fig. 4.
Fig. 4. RIN of MI-FPLD in comparison with that of commercial tunable laser diode and a free-running FPLD.
Fig. 5.
Fig. 5. Effect of amplitude of broadcast signal on downstream signal (a) broadcast signalinduced power penalty (b) RIN as a function of RF power of broadcast signal.
Fig. 6.
Fig. 6. BER curves of (a) downstream data and (b) upstream data while varying the extinction ratio of downstream signal
Fig. 7.
Fig. 7. Measured (a) BER and (b) electrical spectrum after transmission over 20-km SSMF
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.