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

Spectrally efficient next-generation optical access network incorporating a novel CWDM uplink combiner

Open Access Open Access

Abstract

A novel wavelength division multiplexed subcarrier multiplexed (WDM/SCM) broadband (1-Gb/s per user) optical access network (OAN) architecture incorporating a coarse WDM (CWDM) uplink combiner is proposed. The concept is demonstrated through theoretical and experimental validation of a 10 × 1-Gb/s quadurature-phase-shift keying (QPSK) SCM optical network. Low penalty transmission is demonstrated for a proof-of-principle dual channel system with a range of 25 km. In agreement with simulation, experiments show that channel spacings of only 1 GHz are viable for Q factors of more than 18 dB. Simulations indicate that the system will operate with 40 wavelengths, each carrying 10 SCM channels at 1 Gb/s.

©2009 Optical Society of America

1. Introduction

There is currently much interest in the use of optical access networks (OANs) for applications such as fiber to the home, fiber to the curb and fiber to the building [1]. OAN architectures have been demonstrated using different multiplexing schemes including time division multiplexing (TDM), wavelength division multiplexing (WDM), subcarrier multiplexing (SCM) and hybrid WDM/TDM [2-5]. Some of these schemes also use time division multiple access, this enabling many users to share an optical channel [2]. Leveraging recent advances in high speed RF technologies, SCM has been proposed for OAN applications [4]. Moreover, combined with SCM and WDM technologies, WDM/SCM OANs can offer a larger number of end-users for a single fiber compared with the other options. Also, WDM/SCM OANs can be easily overlaid onto current networks by making use of spare fiber and wavelengths. However, there are significant disadvantages which have been acknowledged including the generation of optical beat interference in the uplink when combining multiple sources [6] and poor optical spectral efficiency due to the double side bands of SCM channels. As a solution to these problems, we have previously demonstrated a spectrally efficient multi-user optical network architecture [7]. However, detailed simulations with accurate numerical models of optical components for greater quantitative understanding were not presented.

In this paper, a bandwidth symmetric WDM/SCM OAN architecture to provide large bandwidths (1-Gb/s) to the home is proposed. Here the use of a novel coarse WDM (CWDM) uplink combiner allows the simple combination of high-speed SCM channels and no optical beat interference from multiple local users. In order to show the potential of the technique for a 10 × 1-Gb/s quadrature phase shift keying (QPSK) SCM multi-user optical network, a proof-of-principle uplink system is demonstrated with two users. The performance of the single wavelength system is validated theoretically using a time domain model [8, 9] and used to confirm the potential for 10 channel operation.

2. Network overview

 figure: Fig. 1.

Fig. 1. System schematic showing the downlink and uplink.

Download Full Size | PDF

Figure 1 illustrates the proposed WDM/OAN architecture. The scheme uses multiple WDM wavelengths, each of which carries multiple channels of SCM data in both the downlink and uplink directions. Each wavelength addresses a different group of users. For each WDM wavelength the downlink broadcasts the signal to all recipients in the group, with each user being assigned an RF channel which he/she demodulates on reception. In the uplink, each user’s data is modulated on to an SCM channel on a given source wavelength. The SCM channels, each on a different wavelength are then detected by the optical uplink combiner. The detected signal is then applied within the combiner to a dense WDM (DWDM) source for back-haul to the central office. It should be noted that the optical combiner detects each of the separate SCM channels, as they are all in band, but does not detect the optical beat frequencies of the uplink carriers, as they are too high in frequency.

3. Experimental setup

Figure 2 shows the experimental setup. For a proof-of-principle experiment two SCM-QPSK channels from two users are transmitted over the network. In order to generate the required 1-Gb/s QPSK signals experimentally, two 500-Mb/s “data” and “not data” pseudorandom outputs from an Anritsu MP1763C 12.5GHz pattern generator are de-correlated by propagation along different lengths of cable. They are then used as the in-phase (I) and quadrature (Q) components of the signal, being up-converted onto each orthogonal RF carrier using a broad-band triple balanced mixer and then superimposed to form a QPSK data signal. The two QPSK signals used in the network are then mixed onto different carrier frequencies and then each is used to drive uplink optical source. The two sources have different wavelengths, namely 1560.61 nm and 1563.05 nm respectively, which have sufficiently different wavelengths to avoid optical beat interference. The two optical signals are combined and detected using a 15 GHz bandwidth photodiode. The resulting composite electrical signal is used to re-modulate a third optical source operating at 1562.23 nm which is used to transmit the resulting optical signal containing the two SCM channels over 25 km of standard single mode fiber (SMF), representing the DWDM back-haul to the central office. The back hauled optical signal is then detected with an amplified 15 GHz PIN diode. The resulting electrical signal is amplified and then split and down-converted with each of the original local oscillators. The resulting base band signal is filtered with a 300 MHz Mini-circuit low-pass filter and analyzed with an oscilloscope and an error detector.

 figure: Fig. 2.

Fig. 2. Expermental setup. ONU: optical network unit; CO: central office; LO: local oscillator; PD: photo detector; LD: laser diode.

Download Full Size | PDF

4. Results and discussion

In order to investigate the feasibility of the multichannel 1-Gb/s QPSK SCM OAN, detailed measurements have been made of the Q factor of the down-mixed data as a function of modulation channel spacing or carrier frequency separation as shown in Figure 3(a). It is found that the Q factor of the SCM channel at a subcarrier frequency of 2.7 GHz increases and then stabilizes with increasing channel spacing owing to reduced inter-channel crosstalk. The Q factors of both I and Q signals are 18.90 and 19.12 dB respectively when the spacing between two SCM channels is 1 GHz. The error performance of the uplink is evaluated by measuring the bit error rates obtained for each of the two SCM channels in the uplink as shown in Figure 3(b). The two BER curves with square symbols show back-to-back (BTB) operation with similar sensitivities for each channel and with no error floor. The two BER curves with triangular symbols show that there is only a 1 dB penalty at a BER of 10-9 for transmission over 25 km, representing the DWDM back-haul to the central office.

 figure: Fig. 3.

Fig. 3. (a) Dependence of Q factor on channel spacing and (b) BER measurements.

Download Full Size | PDF

In order to evaluate the possibility of implementing a 10 × 1-Gb/s SCM optical network, a computer simulation using time domain laser model is performed. This simulation includes the performance of the RF components as expected in an integrated system, rather than the discrete components used in our proof-of-principle demonstrator. Figure 4 shows the effect of the 3 dB Gaussian filter bandwidth and the spacing between the SCM channels on the Q factor. The impact of inter-symbol interference (ISI) of a single SCM channel increases with decreasing filter bandwidth, while crosstalk is minimized at the optimum optical filter bandwidth. The optimal filter bandwidth is determined to be 300 MHz when the channel spacing is 1 GHz. Figure 5 demonstrates the feasibility of 10 × 1-Gb/s QPSK SCM transmission spaced at 1 GHz. Here, the subcarrier frequency is chosen to be in the 1.5 GHz to 10.5 GHz frequency range because the first subcarrier frequency of 1.5 GHz is sufficiently high to minimize the effect of the second harmonic distortion on other SCM channels at higher frequencies. The optimized optical modulation index (OMI), defined as the ratio of the sum of maximum & minimum transmitted optical power to the difference of maximum & minimum transmitted optical power, in the optical network unit (ONU) and uplink combiner is 0.49 and 0.97 respectively. The Q factor of an SCM channel is calculated as SCM channels are serially added from low to high frequency. The result shows that the Q factor deteriorates as the channel number gets larger. This is mainly because the power of each SCM channels is reduced with increasing channel number, while the harmonic and inter-modulation distortions of the SCM channels, caused by the laser and Mach-Zehnder (MZ) modulator nonlinearities respectively, are accumulated. In addition, the higher frequency SCM channels are affected by increased chromatic dispersion. However, the Q factor of the worst SCM channels is 19.90 dB even with 10 SCM channels transmitted over 25 km of SMF.

 figure: Fig. 4.

Fig. 4. Dependence of calculated Q factor on ISI and crosstalk as a function of 3 dB Gaussian filter bandwidth.

Download Full Size | PDF

 figure: Fig. 5.

Fig. 5. Q factor of the worst channel versus number of channels for 25 km SMF transmission when an SCM channel is added serially; OMI_ONU: OMI in the ONU and OMI_UC: OMI in the uplink combiner.

Download Full Size | PDF

To determine the optical power budget of a 1 × 10 Gb/s QPSK SCM uplink system, the BER curves of 10 SCM channels as a function of received optical power are firstly calculated as shown in Figure 6. The receiver sensitivity of -19.7 dBm for the worst SCM channel at a 10-9 BER is obtained. Next, the optical power budget is clarified in Table 1. Here, the CWDM transmission loss is ignored due to the very short reach. The proposed uplink system uses one CWDM AWG at the uplink combiner and two DWDM AWGs (one is at the transmitter end in the ONU to combine all the wavelengths and the other is at the receiver end in the central office). Each of them contributes 4 dB loss of power. Fiber attenuation is 0.25 dB/km for the standard SMF and thus 25 km fiber transmission loss is equal to 0.25 × 25 = 6.25 dB. Therefore, the maximum received power is -16.26 dBm and the receiver sensitivity is -19.7 dBm, which results in a power margin of 3.55 dB. These results demonstrate the potential for multi-channel WDM/SCM transmission to achieve a 10-Gb/s aggregate data rate with a 1-Gb/s single user data rate.

 figure: Fig. 6.

Fig. 6. BER calculation in a 10 × 1-Gb/s QPSK SCM uplink system.

Download Full Size | PDF

Tables Icon

Table 1. Power budget calculation in a 10 × 1-Gb/s QPSK SCM uplink system.

5. Conclusion

An architecture for a spectrally efficient WDM/SCM OAN has been proposed in which high bandwidth services are delivered to local users. A CWDM uplink combiner allows the simple combination of local SCM channels from multiple local users. In a proof-of-principle demonstration, a 1-Gb/s QPSK SCM OAN is demonstrated using two uplink channels with a 1 GHz RF carrier separation. A power penalty of 1 dB is found after transmission of 25 km. Simulation results indicate that each optical channel is capable of supporting 10 × 1-Gb/s SCM channels and so potentially serving 400 users simultaneously when using 40 DWDM channels.

Acknowledgment

The authors would like to acknowledge the UK EPSRC funded HIPNet project and financial assistance from the Centre for Advanced Photonics and Electronics, at the University of Cambridge.

References and links

1. H. Shinohara, “Broadband access in Japan: rapidly growing FTTH market,” IEEE J. Sel. Areas Commun . 43, 72–78, (2005).

2. D. P. Shea and J. E. Mitchell, “Operating penalties in single-fiber operation 10-Gb/s, 1024-way split, 110km long-reach optical access networks,” IEEE Photon. Technol. Lett . 18, no 23, 2463–2465, (2006). [CrossRef]  

3. S. Park, C. Lee, K. Jeong, H. Park, J. Ahn, and K. Song, “Fiber-to-the-Home Services Based on Wavelength-Division-Multiplexing Passive Optical Network,” IEEE J. Lightwave Technol . 22, 2582-(2004). [CrossRef]  

4. R. Olshansky, V. A. Lanzisera, and P. M. Hill, “Subcarrier multiplexed lightwave systems for broad-band distribution,” IEEE J. Lightwave Technol . 7, 1329–1342, (1989). [CrossRef]  

5. Dong Jae Shin et al, “Hybrid WDM/TDM-PON with wavelength-selection-free transmitters, IEEE J. Lightwave Technol . 23, no. 1, 187–195, (2005). [CrossRef]  

6. D. C. Desem, “Optical interference in subcarrier multiplexed systems with multiple optical carriers,” IEEE J. Sel. Areas Commun . 8, no. 7, 1290–1295, (1990). [CrossRef]  

7. J. Y. Ha, A. Wonfor, R. V. Penty, I. H. White, and P. Ghiggino, “Spectrally Efficient 10 × 1 Gb/s QPSK Multi-User Optical Network Architecture,” in Proceedings of the Optical Fiber Communication and Conference, paper OWD6 (2007).

8. E. M. Pratt and J. E. Carroll, “Gain modelling and particle balance in semiconductor lasers,” Electron. Lett . 147, 77–82, (2000).

9. G. Smith, D. Novak, and Z. Ahmed, “Overcoming chromatic-dispersion effects in fiber-wireless systems incorporating external modulators,” IEEE Trans. Microwave Theory Tech . 45, 1410–1415, (1997). [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 (6)

Fig. 1.
Fig. 1. System schematic showing the downlink and uplink.
Fig. 2.
Fig. 2. Expermental setup. ONU: optical network unit; CO: central office; LO: local oscillator; PD: photo detector; LD: laser diode.
Fig. 3.
Fig. 3. (a) Dependence of Q factor on channel spacing and (b) BER measurements.
Fig. 4.
Fig. 4. Dependence of calculated Q factor on ISI and crosstalk as a function of 3 dB Gaussian filter bandwidth.
Fig. 5.
Fig. 5. Q factor of the worst channel versus number of channels for 25 km SMF transmission when an SCM channel is added serially; OMI_ONU: OMI in the ONU and OMI_UC: OMI in the uplink combiner.
Fig. 6.
Fig. 6. BER calculation in a 10 × 1-Gb/s QPSK SCM uplink system.

Tables (1)

Tables Icon

Table 1. Power budget calculation in a 10 × 1-Gb/s QPSK SCM uplink system.

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.