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

Simplified flexible-PON upstream transmission using pulse position modulation at ONU and DSP-enabled soft-combining at OLT for adaptive link budgets

Open Access Open Access

Abstract

We demonstrate a flexible modulation and detection scheme for upstream transmission in passive optical networks using pulse position modulation at optical network unit, facilitating burst-mode detection with automatic decision threshold tracking, and DSP-enabled soft-combining at optical line terminal. Adaptive receiver sensitivities of −33.1 dBm, −36.6 dBm and −38.3 dBm at a bit error ratio of 10−4 are respectively achieved for 2.5 Gb/s, 1.25 Gb/s and 625 Mb/s after transmission over a 20-km standard single-mode fiber without any optical amplification.

© 2015 Optical Society of America

1. Introduction

Passive optical networks (PONs) continue to be the subject of active research and development to meet the ever-increasing capacity demand in access networks [1]. Current commercial PON systems are based on binary on-off-keying (OOK) modulation and detection for simplicity and low cost. Digital signal processing (DSP) has recently been used in core networks to enable substantial system performance improvements, and is also being studied for PONs [2–4]. One attractive feature offered by DSP is the ability to vary link loss budget based on link conditions, thereby enabling better utilization of network resources and supporting more users or optical network units (ONUs). This type of DSP-enabled flexible-PON has been demonstrated by using orthogonal frequency division multiplexing (OFDM) with adaptive bit loading [5–7]. When OFDM is used, digital-to-analog converter (DAC) and analog-to-digital converter (ADC) are needed at each optical line terminal (OLT) and in all the ONUs communicating with the OLT. In PONs, the upstream transmission is typically operating in burst mode and is more challenging than the downstream transmission [8]. Here, we propose a simplified flexible-PON upstream transmission scheme based on the combination of pulse-position modulation (PPM) [9,10] at ONUs and DSP-enabled soft-combining [11,12] at OLT. With PPM, the ONU transmitters can remain OOK-like (without DAC and linear modulation), so the ONU can remain as cost-effective as conventional OOK-based ONU essentially. The benefits of PPM include improved receiver sensitivity and automatic decision-threshold tracking for facilitating burst-mode-detection. When combined with DSP-enabled soft-combining at OLT, the link budget can be flexibly adjusted by trading net data rate with receiver sensitivity. One drawback of this scheme is the increased cost of the OLT as compared to conventional OOK-based OLT due to the need for ADC and linear receiver. As each OLT will be shared by many ONUs, the increased OLT cost may only account for a small fraction of the overall PON cost, and may thus be justified in certain applications. We demonstrate this scheme by modulating a 4-ary PPM signal using a directly modulated laser (DML), transmitting it over a 20-km standard single-mode fiber (SSMF), and receiving it with a DSP-enabled receiver with frame synchronization and channel equalization capability. Receiver sensitivities of −33.1 dBm, −36.6 dBm, and −38.3 dBm are obtained at a bit error ratio (BER) of 10−4 without any optical amplification for 2.5 Gb/s, 1.25 Gb/s, and 625 Mb/s, respectively.

2. Experimental setup

Figure 1 shows the schematic of the experimental setup for the generation and detection of a 2.5-Gb/s 4-PPM signal. At the transmitter, a 4-ary PPM signal with 15000 symbols per frame is modulated according to a 215-1 PRBS. A training sequence (TS) containing 16 slots is used at the beginning of each frame for synchronization and channel equalization. The PPM frames are stored in an arbitrary waveform generator (AWG) and repetitively outputted by a DAC in the AWG. Note that the PPM generation only requires binary intensity modulation (IM), as simple as OOK, and we use AWG here just for easy generation of the binary waveform. The DAC sampling rate is first set at 5 GSa/s, corresponding to a slot rate of 5 GHz. For 4-ary PPM, the symbol period (τPPM) is 800 ps and the data rate is 2.5 Gb/s. The DAC output is amplified to a peak-to-peak voltage swing of ~5 V before driving a 1550-nm DML with 3-GHz RF bandwidth. The bias of the DML is controlled to optimize the PPM based on the drive condition. The typical signal output power of the DML is 3 dBm. After 20-km SSMF transmission, a variable optical attenuator (VOA) is used to vary the received optical power (P) before the signal is detected by an avalanche photodiode (APD). The APD is followed by an automatic gain control (AGC) unit with a dynamic range of over 20 dB. The detected signal is then amplified and sampled by a 10-GSa/s ADC in a real-time sampling scope, followed by offline DSP. The key DSP steps include frame synchronization, TS-based channel estimation and compensation, PPM demodulation (by finding the maximum power slot in each PPM symbol), and BER measurement.

 figure: Fig. 1

Fig. 1 Schematic of the experimental setup. DAC: digital-to-analog-converter; DML: directly-modulated laser; VOA: variable optical attenuator; APD: avalanche photodiode; AGC: automatic gain control; ADC: analog-to-digital converter; Sync: synchronization; CE/CC: channel estimation/compensation.

Download Full Size | PDF

Figure 2 (a) shows a representative recovered 4-PPM waveform at the back-to-back configuration, in comparison with the original signal pattern. Clearly, valid signal recovery is obtained. Figure 2(b) shows a digitally recovered signal optical spectrum (at P = −33 dBm) when an intentional DC-wandering is introduced by a particular RF amplifier in order to test the automatic decision-threshold tracking feature. Figure 2(c) shows a recovered signal spectrum at the same received power without the intentional DC-wandering. In both cases, the 5-dB RF bandwidth of the signal is about 2.5 GHz, indicating that it can be generated by low-cost optical components.

 figure: Fig. 2

Fig. 2 (a) A representative recovered 4-PPM waveform and the original signal pattern; (b) Digitally recovered signal spectrum with an intentional DC-wandering (to test automatic decision-threshold tracking); (c) Recovered signal spectrum without the DC-wandering.

Download Full Size | PDF

3. Experimental results

We first test the automatic decision-threshold tracking feature. Figure 3(a) shows a sample recovered PPM frame at P = −33.5 dBm after 20-km SSMF transmission. Clearly, due to the intentional DC-wandering, the PPM waveform largely varies during the frame. Figure 3(b) shows the measured error count per PPM symbol in the recovered frame. There are a few error counts in this frame, and they are distributed over time with reasonable uniformity, indicating that the automatic decision tracking is achieved, by the maximum-likelihood PPM demodulation on a symbol-by-symbol basis. Figure 3(c) shows another sample recovered PPM frame at P = −32 dBm, where no error count is found, as shown in Fig. 3(d), further indicating that automatic decision tracking is functioning properly.

 figure: Fig. 3

Fig. 3 (a) A sample recovered PPM frame at P = −33.5 dBm after 20-km SSMF transmission; (b) Measured error count per PPM symbol in the recovered frame as shown in (a); (c) A sample recovered PPM frame at P = −32 dBm; (d) Measured error count per PPM symbol in the recovered frame as shown in (c).

Download Full Size | PDF

We then compare the performance after 20-km SSMF transmission (L = 20 km) with that at the back-to-back configuration (L = 0 km). As shown in Fig. 4, the fiber dispersion induced power penalty is about 0.5 dB at BER = 10−4, indicating that the 2.5-Gb/s PPM signal is reasonably robust against fiber dispersion. For 10-Gb/s PON, the upstream signal wavelength band is around 1.3 μm, where the dispersion of SSMF is much less than that at 1550 nm, so dispersion-induced penalty there will be negligible. The receiver sensitivity at BER = 10−4 is −33.1 dBm, which meets the second most stringent sensitivity requirement specified in the 10G-PON standards [13], E1 (−31.5 dBm at BER = 10−4). This is notable as the E1 requirement is originally intended for optically amplified PON, while we meet the E1 requirement without optical amplification at both the OLT and the ONU.

 figure: Fig. 4

Fig. 4 Measured BER versus received optical power at L = 0 km and L = 20 km.

Download Full Size | PDF

To support even higher loss budget via rate adaptation, we first reduce the 4-PPM slot rate by a factor of two (to 2.5 GHz), so the data rate is reduced to 1.25 Gb/s. To further increase loss budget, we may further reduce the slot rate, but this may also reduce the decision tracking speed as each PPM symbol will be longer in time. We resort to repetition coding with soft combining to provide additional rate adaptation [11]. Figure 5 shows the measured BER after 20-km SSMF transmission as a function of the received optical power for three cases. The first case is the original 2.5-Gb/s 4-PPM with 5-GHz slot rate, the second case is 1.25-Gb/s 4-PPM with 2.5-GHz slot rate, and the third case is 625-MHz 4-PPM with 2.5-GHz slot rate and soft-combing of two repetitive copies of the signal. For the repetition coding, the repetition is made to occur in adjacent PPM symbols. At the typical BER threshold of 10−4 [13], receiver sensitivities of −33.1 dBm, −36.6 dBm, and −38.3 dBm are obtained at 2.5 Gb/s, 1.25 Gb/s, and 6.25 Mb/s, respectively. Thus, the most stringent sensitivity requirement specified in the 10G-PON standards [13], E2 (−33.5 dBm at BER = 10−4), can be achieved in the two latter cases with 3.1 dB and 4.8 dB margins, respectively. Note that the improvements in receiver sensitivities or loss budgets are obtained at the expense of reduced data rates. This may be a worthy tradeoff for certain applications where increased loss budget is desired.

 figure: Fig. 5

Fig. 5 Measured BER after 20-km SSMF vs. the received power (P) for three different data rates, 2.5 Gb/s, 1.25 Gb/s, and 625 Mb/s, respectively corresponding to 4-PPM with 5-GHz slot rate, 2.5-GHz slot rate, and 2.5-GHz slot rate and soft-combining of two repeated copies.

Download Full Size | PDF

To visualize the performance improvement enabled by the rate adaptation techniques, we show in Fig. 6 sample recovered signal power waveforms after 20-km SSMF transmission for three different data rates, 2.5 Gb/s, 1.25 Gb/s, and 625 Mb/s, respectively corresponding to 4-PPM with 5-GHz slot rate, 2.5-GHz slot rate, and 2.5-GHz slot rate and soft-combining of two repeated copies. The received power level in these three cases is −35 dBm, and the intentional DC-wandering is turned off in order to better show the differences among these three cases. At 2.5 Gb/s, the pulses in the PPM signal waveform are intermixed with the zeros, as shown in Fig. 6(a), and the measured BER is 6.1 × 10−3. When the signal data rate is reduced to 1.25 Gb/s via PPM slot rate reduction, the pulses in the PPM signal waveform are nearly separated from the zeros, as shown in Fig. 6(b), and no errors are recorded with over 106 bits measured. When the signal data rate is further reduced to 625 Mb/s via soft-combining with two repeated copies, the pulses in the PPM signal waveform are well separated from the zeros, as shown in Fig. 6(c), and no errors are recorded with over 106 bits measured. Given the DML output power of 3 dBm, the link loss budgets achieved at BER = 10−4 are thus 36.1, 39.6 and 41.3 dB for 2.5 Gb/s, 1.25 Gb/s, and 625 Mb/s, respectively. The achieved loss budgets all meet the most stringent loss budget requirement specified in the 10G-PON standards [13]. With the use of optical amplification at the OLT, e.g., via a compact semiconductor optical amplifier [14], the achievable link budgets are expected to be further increased.

 figure: Fig. 6

Fig. 6 Sample recovered signal power waveforms measured at P = −35 dBm for three different data rates, 2.5 Gb/s (a), 1.25 Gb/s (b), and 625 Mb/s (c), respectively corresponding to 4-PPM with 5-GHz slot rate, 2.5-GHz slot rate, and 2.5-GHz slot rate and soft-combining of two repeated copies.

Download Full Size | PDF

4. Conclusion

We have proposed and experimentally demonstrated a novel flexible-PON upstream transmission scheme using simple binary-IM based pulse-position modulation at the ONUs, and a DSP-enabled soft-combining receiver at the OLT. Automatic decision-threshold tracking has been achieved by using maximum-likelihood PPM demodulation on a symbol-by-symbol basis. Adaptive link loss budgets have been realized by PPM slot rate reduction and repetition coding with soft combining. Receiver sensitivities of −33.1 dBm, −36.6 dBm, and −38.3 dBm at a BER of 10−4 have been respectively obtained for signal data rates of 2.5 Gb/s, 1.25 Gb/s, and 625 Mb/s after 20-km SSMF transmission without any optical amplification. This type of simplified flexible-PON architecture may be a viable practical solution for future PON systems, where cost-sensitive ONUs remain essentially as simple as current OOK-based ONUs while more sophisticated DSP-based transmitters and/or receivers are used in OLTs to offer increased flexibility and overall cost advantage.

References and links

1. D. Nesset, “NG-PON2 technology and standards,” in Proc. European Conference on Optical Communications (ECOC) (2014), tutorial paper Mo.4.1.1.

2. N. Cvijetic, M. Cvijetic, M.-F. Huang, E. Ip, Y.-K. Huang, and T. Wang, “Terabit optical access networks based on WDM-OFDMA-PON,” J. Lightwave Technol. 30(4), 493–503 (2012). [CrossRef]  

3. N. Yoshimoto, J. Kani, S.-Y. Kim, N. Iiyama, and J. Terada, “DSP-based optical access approaches for enhancing NG-PON2 systems,” IEEE Commun. Mag. 51(3), 58–64 (2013). [CrossRef]  

4. D. Lavery and S. J. Savory, “Digital coherent technology for long-reach optical access,” in Proc. Optical Fiber Communications Conference (OFC) (2014), invited paper Tu2F.1. [CrossRef]  

5. H. Kimura, K. Asaka, H. Nakamura, S. Kimura, and N. Yoshimoto, “First demonstration of energy efficient IM-DD OFDM-PON using dynamic SNR management and adaptive modulation using dynamic SNR management and adaptive modulation,” in Proc. European Conference on Optical Communications (ECOC) (2013), paper We.4.F.5.

6. R. Giddings, “Real-time digital signal processing for future optical access networks,” in Proc. Optical Fiber Communications Conference (OFC) (2013), tutorial paper OM3H.5. [CrossRef]  

7. L. Zhou, N. Chand, X. Liu, G. Peng, H. Lin, Z. Li, Z. Wang, X. Zhang, S. Wang, and F. Effenberger, “Demonstration of software-defined flexible-PON with adaptive data rates between 13.8 Gb/s and 5.2 Gb/s supporting link loss budgets between 15 dB and 35 dB,” in Proc. European Conference on Optical Communications (ECOC) (2014), paper P.7.24. [CrossRef]  

8. X. Z. Qiu, “Burst-mode receiver technology for short synchronization,” in Proc. Optical Fiber Communications Conference (OFC) (2013), tutorial paper OW3G.4.

9. D. O. Caplan, B. S. Robinson, R. J. Murphy, and M. L. Stevens, “Demonstration of 2.5 Gslot/sec optically-preamplified M-PPM with 4 photons/bit receiver sensitivity,” in Proc. Optical Fiber Communications Conference (OFC) (2005), post-deadline paper PDP32.

10. X. Liu, T. H. Wood, R. W. Tkach, and S. Chandrasekhar, “Demonstration of record sensitivity in an optically pre-amplified receiver by combining PDM-QPSK and 16-PPM with pilot-assisted digital coherent detection,” in Proc. Optical Fiber Communications Conference (OFC) (2011), post-deadline paper PDB1. [CrossRef]  

11. G. Gho, L. Klak, and J. M. Kahn, “Rate-adaptive coding for optical fiber transmission systems,” J. Lightwave Technol. 29(2), 222–233 (2011). [CrossRef]  

12. X. Liu, “Twin-wave transmission with enhanced performance,” in Proc. European Conference on Optical Communications (ECOC) (2014), invited paper We.2.3.3.

13. ITU-T Recommendation G.987.2: 10-Gigabit-capable passive optical networks (XG-PON): Physical media dependent (PMD) layer specification (2010).

14. R. Bonk, H. Schmuck, W. Poehlmann, and T. Pfeiffer, “Beneficial OLT transmitter and receiver concepts for NG-PON2 using semiconductor optical amplifiers,” in Proc. Optical Fiber Communications Conference (OFC) (2014), paper W1D.5. [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 Schematic of the experimental setup. DAC: digital-to-analog-converter; DML: directly-modulated laser; VOA: variable optical attenuator; APD: avalanche photodiode; AGC: automatic gain control; ADC: analog-to-digital converter; Sync: synchronization; CE/CC: channel estimation/compensation.
Fig. 2
Fig. 2 (a) A representative recovered 4-PPM waveform and the original signal pattern; (b) Digitally recovered signal spectrum with an intentional DC-wandering (to test automatic decision-threshold tracking); (c) Recovered signal spectrum without the DC-wandering.
Fig. 3
Fig. 3 (a) A sample recovered PPM frame at P = −33.5 dBm after 20-km SSMF transmission; (b) Measured error count per PPM symbol in the recovered frame as shown in (a); (c) A sample recovered PPM frame at P = −32 dBm; (d) Measured error count per PPM symbol in the recovered frame as shown in (c).
Fig. 4
Fig. 4 Measured BER versus received optical power at L = 0 km and L = 20 km.
Fig. 5
Fig. 5 Measured BER after 20-km SSMF vs. the received power (P) for three different data rates, 2.5 Gb/s, 1.25 Gb/s, and 625 Mb/s, respectively corresponding to 4-PPM with 5-GHz slot rate, 2.5-GHz slot rate, and 2.5-GHz slot rate and soft-combining of two repeated copies.
Fig. 6
Fig. 6 Sample recovered signal power waveforms measured at P = −35 dBm for three different data rates, 2.5 Gb/s (a), 1.25 Gb/s (b), and 625 Mb/s (c), respectively corresponding to 4-PPM with 5-GHz slot rate, 2.5-GHz slot rate, and 2.5-GHz slot rate and soft-combining of two repeated copies.
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.