First demonstration of a scalable MDM/CDM optical access system

Takahiro Kodama; Tomoki Isoda; Koji Morita; Akihiro Maruta; Ryo Maruyama; Nobuo Kuwaki; Shoichiro Matsuo; Naoya Wada; Gabriella Cincotti; Ken-ichi Kitayama

doi:10.1364/OE.22.012060

1. Introduction

The increasing demand for broadband and multimedia services boosts to upgrade current time division multiplexing (TDM)-based optical access systems toward the next generation-passive optical network (PON) stage-2 (NG-PON2). Benefits expected are related to larger transmission capacity, increased geographic coverage, scaling of the number of optical node units (ONU), with a reduction of the number of central offices (CO), wireless and wireline integration, and an overall reduction of capital expenditure (CAPEX) and operating expenditure (OPEX).

Time and wavelength division multiplexing (TWDM) technology has been selected by the Full Service Access Network (FSAN) group as a primary solution for NG-PON2 systems, mainly to the ability to operate on existing optical distribution networks (ODN) and the compatibility with video overlay. To stack four 10G-PONs stage-1 (XG-PON1) using four pairs of wavelengths, a TWDM system requires tunable transceivers in the CO and ONUs; in addition, the coexistence with legacy PONs strictly depends on the wavelength plan.

Beyond the wavelength domain, there are other possible ways to efficiency increase the network capacity, and deal with the fiber capacity crunch [1]. Nowadays, space-division multiplexing (SDM) and mode-division multiplexing (MDM) are attracting a lot of research interest, and several architectures for mode multiplexer/demultiplexer (MMUX/MDeMUX) have been proposed in literature [2–4]. Free-space optics devices have been fabricated [5], and MDM transmission of 10 Gbps, non-return-to-zero (NRZ)-on-off-keying (OOK) signals through 2 km-long two mode fiber (TMF) with low mode coupling have experimentally demonstrated [6, 7]. In addition, low-loss cost-effective integrated Silicon-on-Insulator [8] and planar lightwave circuit (PLC) [9] MMUX/MDeMUX have been developed for MDM-based PONs, and C-band wavelength division multiplexing (WDM)-MDM transmission has been also demonstrated [10].

A further capacity increase, with a larger flexibility and bandwidth granularity, can be achieved using hybrid system, combining MDM with code division multiplexing (CDM) [11]; the latter approach provides another dimension for asynchronous multiple access, beyond space, time and frequency, allowing bit-rate symmetric and fully asynchronous transmission, that are the key features of NG-PON3 systems [12]. Table 1 reports a comparison between WDM- and MDM-based technologies [13–15]. It is evident that MDM systems are energy efficient and cost-effective because they do not require large, or tunable, broadband light sources, such as super continuum, that are expensive and power consuming. In addition, they require only commercial erbium-doped fiber amplifiers (EDFA), with a constant gain in a few nm bandwidth.

Table 1. Comparison between WDM-CDM and MDM-CDM.

View Table | View all tables in this article

During the last decade, most part of our research has been devoted to the study of CDM-based PON systems. We have experimentally demonstrated long-reach dispersion-compensation-free CDM transmission [16], and to satisfy the colorless (non-user specific) condition for ONUs [17], we adopted PON configurations with paired multiport encoder/decoders (E/D), located at the remote node (RN) and in the CO. In addition, to fully exploit the optical resources (i.e. the coding/decoding capability of E/D), a cascaded E/D configuration for multiple noise suppression at RN has been proposed [18].

In the present paper, we demonstrate that, beyond conventional WDM approach, MDM and CDM are suitable technologies for NG-PON3 systems and we propose a novel cost-effective all-optical hybrid MDM-CDM system with a MMUX/MDeMUX at the RN and CO, using a single wavelength. We successfully demonstrate asynchronous OOK modulation, 2 mode x 4 code x 10 Gbps uplink transmission, over a 40-km single mode fiber (SMF) and a 2-km TMF, without dispersion compensation, considering three different geographic conditions. Chromatic and mode dispersion impairments have been investigated by measuring the bit-error-rate (BER). To analyze the mode crosstalk effect due to the power difference between the LP₀₁ and LP₁₁ modes, we measure the BER performance considering a 30 km span length difference. Finally, we also evaluated the maximum mode crosstalk tolerance due to a function of the power difference between the two access spans.

2. System architecture

Figure 1 shows the MDM-CDM system architecture, with a pair of optical MMUX/MDeMUX located at the passive RN and at the CO, where two Optical Line Terminals (OLT) are placed. It is evident that the proposed architecture legacy PON systems; on the other hand, the introduction of a MMUX/MDeMUX in the CO, allows SMF-based optical signal processing. We consider the 40 km minimum access span length of the NG-PON2 standard. The LP₀₁ and LP₁₁ modes are generated at the same MMUX output port, the choice of the input port determines which mode is created. The MDeMUX has the same configuration as the MMUX [5, 9, 10], and each mode is detected at the corresponding output port.

Fig. 1 MDM/CDM system using a TMF.

Download Full Size | PDF

A multiport E/D and a narrowband optical band-pass filter (NB-OBPF) are placed both at the OLT and at the sub-RNs, to optically encode/decode the signals. The PLC-based E/D has the unique capability of simultaneously generating and processing multiple phase-shift-keying (PSK) OCs [19].

The NB-OBPF is used to extend the reach length, without dispersion compensation. In addition, the RN and sub-RNs are on the transmission line between the CO and the ONUs, so that the splitting losses are limited; this configuration is similar to a conventional WDM-based PON, where an arrayed waveguide grading (AWG) is located in the RN [20].

The proposed system is able to reduce the number of COs, gather OLTs together and directly connect metro/core networks, for cost-effective high-capacity NG-PON3 systems. This technology can be also used for unbundling the local loop (ULL) for open-access optical networks.

The same OC can be used for different modes, so that a CDM-based 10G-PON can be scaled up by a factor equal to the number of modes. For the case of two modes and 16-port E/Ds, the maximum number of ONUs is 32.

3. Numerical model

To analyze the performance of a two-mode OOK, MDM-CDM system using data rate detection, we consider a complete model, taking into account the crosscorrelation signals due to the unmatched codes, the mode crosstalk (desired and undesired modes), and the beat noise at the receiver, as it is shown in Fig. 2.The received signal can be expressed as

\begin{array}{l} Z = ℜ \int_{0}^{T_{b}} E_{d - a c}^{2} (t) d t + ℜ \sum_{i = 2}^{N} \int_{0}^{T_{b}} E_{d - c c - i}^{2} (t) d t + ℜ \int_{0}^{T_{b}} E_{u - a c}^{2} (t) d t + ℜ \sum_{i = 2}^{N} \int_{0}^{T_{b}} E_{u - c c - i}^{2} (t) d t \\ + 2 ℜ \sum_{i = 2}^{N} \int_{0}^{T_{b}} | E_{d - a c} (t) | | E_{d - c c - i} (t) | \cos [\emptyset_{d - a c} (t) - \emptyset_{d - c c - i} (t)] d t \\ + 2 ℜ \int_{0}^{T_{b}} | E_{d - a c} (t) | | E_{u - a c} (t) | \cos [\emptyset_{d - a c} (t) - \emptyset_{u - a c} (t)] d t \\ + 2 ℜ \sum_{i = 2}^{N} \int_{0}^{T_{b}} | E_{d - a c} (t) | | E_{u - c c - i} (t) | \cos [\emptyset_{d - a c} (t) - \emptyset_{u - c c - i} (t)] d t \\ + 2 ℜ \sum_{i, j = 2 (i > j)}^{N} \int_{0}^{T_{b}} | E_{d - c c - i} (t) | | E_{d - c c - j} (t) | \cos [\emptyset_{d - c c - i} (t) - \emptyset_{d - c c - j} (t)] d t \\ + 2 ℜ \sum_{i = 2}^{N} \int_{0}^{T_{b}} | E_{d - c c - i} (t) | | E_{u - a c} (t) | \cos [\emptyset_{d - c c - i} (t) - \emptyset_{u - a c} (t)] d t \\ + 2 ℜ \sum_{i, j = 2 (i > j)}^{N} \int_{0}^{T_{b}} | E_{d - c c - i} (t) | | E_{u - c c - j} (t) | \cos [\emptyset_{d - c c - i} (t) - \emptyset_{u - c c - j} (t)] d t \\ + 2 ℜ \sum_{i = 2}^{N} \int_{0}^{T_{b}} | E_{u - a c} (t) | | E_{u - c c - i} (t) | \cos [\emptyset_{u - a c} (t) - \emptyset_{u - c c - i} (t)] d t \\ + 2 ℜ \sum_{i, j = 2 (i > j)}^{N} \int_{0}^{T_{b}} | E_{u - c c - i} (t) | | E_{u - c c - j} (t) | \cos [\emptyset_{u - c c - i} (t) - \emptyset_{u - c c - j} (t)] d t + n (t), \end{array}

where E_d-ac exp(φ_d-ac) is the autocorrelation of the desired mode, E_d-ccexp(φ_d-cc), E_u-acexp(φ_u-ac), and E_u-ccexp(φ_u-cc) are the crosscorrelations of the desired mode, the autocorrelation of the undesired mode, and the crosscorrelations of the undesired mode, respectively.

ℜ

is the photodetector (PD) responsivity and T_b the symbol duration. Therefore, the first term represents the matched signal, the second, third and fourth terms are the multiple access interference noises, the fifth, sixth and seventh terms are the first-order beat noises, from the eighth to twelfth terms are the second-order beat noises. The last term is a Gaussian random noise, related to thermal and shot noises. When the code performance is poor, and the mode extinction ratio is high (|E_d-cc| >> |E_u-ac|), the crosscorrelations (second term) and the beat noise due to crosscorrelations (fifth term) are dominant. On the other hand, when the codes are orthogonal to each other (and the crosscorrelation signals vanish), and the power difference between the modes is large (|E_d-cc|<<|E_u-ac|), the mode crosstalk (third term) and the beat noise due to mode crosstalk (sixth term) have the foremost detrimental effects. However, in the following experiments, we have simplified the system model considering only multiple access noises (crosscorrelation signals), mode crosstalk and beat noises related to autocorrelation of the desired mode. The detected signal can be written as

Fig. 2 Mode crosstalk evaluation model.

Download Full Size | PDF

\begin{array}{l} Z = ℜ \int_{0}^{T_{b}} E_{d - a c}^{2} (t) d t + ℜ \sum_{i = 2}^{N} \int_{0}^{T_{b}} E_{d - c c - i}^{2} (t) d t + ℜ \int_{0}^{T_{b}} E_{u - a c}^{2} (t) d t \\ + 2 ℜ \sum_{i = 2}^{N} \int_{0}^{T_{b}} | E_{d - a c} (t) | | E_{d - c c - i} (t) | \cos [\emptyset_{d - a c} (t) - \emptyset_{d - c c - i} (t)] d t \\ + 2 ℜ \int_{0}^{T_{b}} | E_{d - a c} (t) | | E_{u - a c} (t) | \cos [\emptyset_{d - a c} (t) - \emptyset_{u - a c} (t)] d t + n (t) . \end{array}

4. Experiments

4.1 2 mode x 4 code uplink transmission

Figure 3 shows the experimental setup and results of an asynchronous, uplink dispersion-compensation-free 2-MDM x 4-CDM-based PON system. We consider three different geographic conditions, so that the two modes propagate over different distances, as shown in Table 2; however, all the time waveforms and spectra have been measured for the case A. The TMF is 2-km long, and the total chromatic dispersions for the two modes at 1550 nm are 43.8 (LP₀₁) and 21.2 (LP₁₁) ps/km/nm, respectively. A NB-OBPF is placed at the sub-RN to tailor the signal spectrum, enabling the dispersion-compensation-free transmission. The MMUX and MDeMUX placed in the RN consist of a suitably designed phase plate and a beam splitter; the MMUX has the peculiar feature of no wavelength dependence over the entire band. Two multiport E/Ds are located at the sub-RN and at the OLT, as shown in Fig. 3.

Fig. 3 Experimental setup for the ONUs and the sub-RN.

Download Full Size | PDF

Table 2. Three different cases for span distances.

View Table | View all tables in this article

A mode-locked laser diode (MLLD) with 9.95328 GHz repetition rate at 1550 nm central wavelength is used to generate a 2.4ps optical pulse stream, modulated by a 2⁷-1 pseudo random bit sequence (PRBS) by a Lithium Niobate-intensity-modulator (LN-IM), as shown in Fig. 3(a). At the sub-RN, four PSK OCs are generated at the E/D output ports #1, #5, #9 and #13, as shown in Fig. 3(b), that are combined together by a set of couplers. Figure 3(c) shows the 4 OC multiplexed signal. Each encoded signal has equal power, random bit phase and same polarization, because we are interested in the worst case analysis.

The 4 OC signals are multiplexed together and transmitted into the SMF#A; to generate the two sets of CDM signals, we use a 3dB coupler.

At the RN, the two CDM signals are sent into the MMUX. One of the LP₀₁ modes is converted into the LP₁₁ mode, by passing through a specially designed phase plate, where half mode phase is inverted. The LP₁₁ and LP₀₁ modes are multiplexed together and sent into the TMF. The 2 modes x 4 OCs multiplexed signal is transmitted through a 2km-long TMF.

At the OLT, the LP₁₁ mode is converted back to LP₀₁ mode by a reversed MMUX, as shown in Figs. 4(a) and 4(b). Although the hybrid system allows asynchronous access to all the users, to analyze the system performance limit, we consider the worst case condition of synchronous transmission, when the beat noise has the most detrimental effect, because the crosscorrelation peaks overlap with the autocorrelation peak, as shown in Fig. 2. In a previous work, we have analyzed the BER performance in the best and worst cases, by changing the interference effect related to the codes arrival time difference [18]. In a fully asynchronous transmission, BER values range between these upperbound and lowerbound limits, and a proper threshold level should be selected.

Fig. 4 Experimental setup of the RN and OLT.

Download Full Size | PDF

We observe that the eye diagram of Fig. 4(a) is not as clear as that one of Fig. 4(b), because the LP₁₁ mode is strongly interfered by the LP₀₁ mode. The measured transmission losses of the LP₀₁ and LP₁₁ modes are 8.5 and 11.8 dB, respectively, due to the half-mirror insertion. The crosstalk between LP₀₁ and LP₁₁ modes, resulting from the MMUX and MDeMUX is approximately −18 dB. Each mode demultiplexed signal is launched into the multiport E/D to obtain the autocorrelation signals; a clear eye opening is obtained after decoding as shown in Figs. 4(c) and 4(d), and the measured code crosscorrelation is −15.5 dB. Therefore, the both code crosscorrelation and mode crosstalk affect the system performance. The decoded signals can be expressed as in Eq. (2) and regenerated into an electrical data pattern by a PD.The BER is measured using a BER tester, after adjusting the decoded signal power at the E/D output.

Figure 5 shows the measured BER for the LP₀₁ and LP₁₁ modes; the dashed, dot, and solid lines refer to the case A, B, and C, respectively; in all cases, BER less than 10⁻³ has been achieved. This confirms error-free transmission, with the aid of a 7%-overhead forward error correction (FEC) of Reed-Solomon (RS) (255, 233). In the case B, the LP₀₁ mode performance worsens by increasing the mode crosstalk, but the MDM-CDM signal can still be error-free received. In the case C, the power penalties increase, due to the cumulative chromatic dispersion effect, as well as the optical signal to noise ratio (OSNR) degradation.

Fig. 5 BER for (a) LP₀₁ mode, (b) LP₁₁ mode.

Download Full Size | PDF

4.2 Mode crosstalk tolerance of access span difference

To evaluate the maximum mode crosstalk tolerance due to different access span lengths, we change the powers of the LP₀₁ and LP₁₁ modes; Fig. 6 reports the experimental setup, and we consider a single OC, because the crosscorrelation terms are negligible if |E_d-cc|<<|E_u-ac|. The decision signal can be expressed as

Z = ℜ \int_{0}^{T_{b}} E_{d - a c}^{2} (t) d t + ℜ \int_{0}^{T_{b}} E_{u - a c}^{2} (t) d t + 2 ℜ \int_{0}^{T_{b}} | E_{d - a c} (t) | | E_{u - a c} (t) | \cos [\emptyset_{d - a c} (t) - \emptyset_{u - a c} (t)] d t + n (t) .

A 3dB coupler splits the data-modulated and encoded signal and the optical power of the two signals is varied using an EDFA, followed by an attenuator (ATT). The two signals are mode multiplexed by the MMUX and launched into a 2km-long TMF. By adjusting the relative power difference between the two modes, both mode distribution profiles are clearly detected by a camera, as shown in Fig. 6. After propagation, the signal mode is demultiplexed, decoded, and photodetected; the BER is measured after amplification. The average power of the undesired mode is set to 0 dBm and the power of the other mode is varied. Figures 6(a)-6(c) and 6(d)-6(f) show the LP₁₁ and LP₀₁ decoded signals, in presence of mode crosstalk, when the power difference between the two modes is 6, 8, and 10 dB. Figure 7 shows the measured BER for each mode. The experimental results show that the OOK, 10Gbps MDM-CDM signal can be error-free transmitted (BER < 10⁻³), if the two modes power difference is lower than 10 dB.

Fig. 6 Experimental setup to evaluate the mode crosstalk performance.

Download Full Size | PDF

Fig. 7 Power difference as a function of BER.

Download Full Size | PDF

The power penalties of each mode, i.e. the variation of the receiver sensitivity for BER = 10⁻⁶ are 1.3, 1.4, 1 and 1.3 dB, due to the insertion losses of the half mirror in the MMUX/MDeMUX. Therefore, we conclude that, neglecting dispersion effects and considering 0.2 dB/km SMF loss, the proposed architecture allows us to place ONUs at distances from the CO, with span length difference lower than 50 km.

5. Conclusions

We have proposed an innovative, flexible and energy-efficient MDM-CDM architecture for future NG-PON3 systems. The proposed approach can be used to enhance the capacity and the number of ONUs in legacy PONs, reducing the number of COs, and directly connect metro/core networks.

We successfully demonstrate an uplink asynchronous transmission over 42 km using a set of SMF and TMF, without dispersion compensation. Since the most of mode crosstalk is generated at the MMUX/MDeMUX, it is possible to extend the TMF transmission length over 2 km, considering only dispersion effect, and this will be the topic of a future work. Moreover, BER performance has been experimentally evaluated, considering mode crosstalk tolerance. The experimental results show that a MDM-CDM signal is still acceptable with a 10 dB power difference between the LP₀₁ and LP₁₁ modes, that corresponds to ONUs located with a 50-km different span distances from the CO.

Acknowledgments

This work will be presented in part at OFC2014, San Francisco, CA, March 2014. The research is supported under the Japan Society for the Promotion of Science (JSPS). This work was partly supported by the NICT R&D program “Basic Technologies for High-Performance Opto-Electronic Hybrid Packet Router” (2011-2016), the European Commission through the ICT-ASTRON project (Contract No. 318714) funded under the 7th Framework Programme, and the Italian Ministry of University and Research through the ROAD-NGN project (PRIN2010-2011).

References and links

1. D. J. Richardson, J. M. Fini, and L. E. Nelson, “Space-division multiplexing in optical fibers,” Nat. Photonics 7(5), 354–362 (2013). [CrossRef]

2. E. Ip, N. Bai, Y. K. Huang, E. Mateo, F. Yaman, M. J. Li, S. Bickham, S. Ten, Y. Luo, G. D. Peng, G. Li, T. Wang, J. Li, C. Montero, and V. Moreno, “6x6 MIMO transmission over 50+25+10km heterogeneous spans of few-mode fiber with inline erbium-doped fiber amplifier,” OFC NFOEC2012, OTu2C (2012).

3. H. Chen, V. Sleiffer, B. Snyder, M. Kuschnerov, R. V. Uden, Y. Jung, C. M. Okonkwo, O. Raz, P. O’Brien, H. de Waardt, and T. Koonen, “Demonstration of a photonic integrated mode coupler with MDM and WDM transmission,” IEEE Photon. Technol. Lett. 25(21), 2039–2042 (2013). [CrossRef]

4. A. Li, X. Chen, A. A. Amin, and W. Shieh, “Fused fiber mode couplers for few-mode transmission,” IEEE Photon. Technol. Lett. 24(21), 1953–1956 (2012). [CrossRef]

5. R. Ryf, S. Randel, A. H. Gnauck, C. Bolle, A. Sierra, S. Mumtaz, M. Esmaeelpour, E. C. Burrows, R.-J. Essimbre, P. J. Winzer, D. W. Peckham, A. H. McCurdy, and R. Lingle, “Mode-division multiplexing over 96km of few-mode fiber using coherent 6x6 MIMO processing,” IEEE/OSA J. Lightwave Technol. 30(4), 521–531 (2012). [CrossRef]

6. K. Kitayama, Y. Kato, S. Seikai, N. Uchida, M. Akiyama, and O. Fukuda, “Transmission characteristic measurement of two-mode optical fiber with a nearly optimum indexprofile,” IEEE Trans. Microw. Theory Tech. 28(6), 604–608 (1980). [CrossRef]

7. T. Isoda, H. Terauchi, K. Tatsumi, A. Maruta, R. Maruyama, N. Kuwaki, S. Matsuo, and K. Kitayama, “Mode division multiplexed transmission through two-mode fiber using space-optics based mode multiplexer/demultiplexer,” in 201318th OptoElectronics and Communications Conference held jointly with 2013 International Conference on Photonics in Switching (Optical Society of America, 2013), paper TuS4_3.

8. H. Chen, R. V. Uden, C. M. Okonkwo, B. Snyder, O. Raz, P. O’Brien, H. V. D. Boom, H. D. Waardt, and T. Koonen, “Employing an integrated mode multiplexer on silicon-on-insulator for few-mode fiber transmission,” Optical Communication (ECOC 2013), 39th European Conference and Exhibition on, 2-26 Sept. 2013, paper Tu.1.B.4.

9. M. Oguma, T. Kitoh, A. Mori, and H. Takahashi, “Ultrawide-passband tandem MZI-synchronized AWG and group delay ripple balancing out technique,” Optical Communication (ECOC), 2010 36th European Conference and Exhibition on, 19-23 Sept. 2010, We.8.E.

10. N. Hanzawa, K. Saitoh, T. Sakamoto, T. Matsui, K. Tsujikawa, M. Koshiba, and F. Yamamoto, “Two-mode PLC-based mode multi/demultiplexer for mode and wavelength division multiplexed transmission,” Opt. Express 21(22), 25752–25760 (2013). [CrossRef] [PubMed]

11. J. A. Salehi, “Code division multiple-access techniques in optical fiber networks. I. fundamental principles,” IEEE Trans. Commun. 37(8), 824–833 (1989). [CrossRef]

12. H. Nakamura, “NG-PON2 Technology,” Optical Fiber Communication Conference and Exposition and the National Fiber Optic Engineers Conference (OFC/NFOEC), (2013), paper Nth4F.5.

13. P. C. Teh, M. Ibsen, J. H. Lee, P. Petropoulos, and D. J. Richardson, “Demonstration of a four-channel WDM/OCDMA system using 255-chip 320-Gchip/s quarternary phase coding gratings,” IEEE Photon. Technol. Lett. 14(2), 227–229 (2002). [CrossRef]

14. X. Wang, N. Wada, N. Kataoka, T. Miyazaki, G. Cincotti, and K. Kitayama, “100km field trial of 1.24 Tbit/s, spectral efficient, asynchronous 5 WDM x 25 DPSK-OCDMA using one set of 50 x 50 ports large scale en/decoder,” in Optical Fiber Communication Conference and Exposition and The National Fiber Optic Engineers Conference, OSA Technical Digest Series (CD) (Optical Society of America, 2007), paper PDP14.

15. H. Beyranvand and J. A. Salehi, “Efficient optical resource allocation and QoS differentiation in optical burst switching networks utilizing hybrid WDM/OCDM,” IEEE/OSA J. Lightwave Technol. 30(15), 2427–2441 (2012). [CrossRef]

16. N. Kataoka, N. Wada, X. Wang, G. Cincotti, and K. Kitayama, “Demonstration of 10Gbps, 4-user, OCDMA transmission over 59km single mode fiber without inline dispersion compensation,” Optical Fiber Communication (OFC), collocated National Fiber Optic Engineers Conference, 2010 Conference on (OFC/NFOEC) 21-25 March 2010.

17. G. Cincotti, N. Kataoka, N. Wada, X. Wang, T. Miyazaki, and K. Kitayama, “Demonstration of asynchronous, 10Gbps OCDMA PON system with colorless and sourceless ONUs,” Optical Communication, 2009. ECOC '09. 35th European Conference on, 20-24 Sept. 2009, 6.5.7.

18. T. Kodama, N. Wada, G. Cincotti, and K. Kitayama, “Asynchronous OCDM-based 10G-PON using cascaded multiport E/Ds to suppress MAI noise,” IEEE/OSA J. Lightwave Technol. 31(20), 3258–3266 (2013). [CrossRef]

19. G. Cincotti, N. Wada, and K. Kitayama, “Characterization of a full encoder/decoder in the AWG configuration for code-based photonic routers. Part I: modeling and design,” IEEE/OSA J. Lightwave Technol. 24(1), 103–112 (2006). [CrossRef]

20. H. Nakamura, K. Taguchi, S. Tamaki, T. Mizuno, Y. Hashizume, T. Yamada, M. Ito, H. Takahashi, S. Kimura, and N. Yoshimoto, “40Gbit/s-class-λ-tunable WDM/TDM-PON using λ-selectable B-Tx and 4 x M cyclic AWG router for flexible photonic aggregation networks,” Opt. Express 21(1), 463–468 (2013). [CrossRef] [PubMed]

	WDM-CDM	MDM-CDM
Light source	Large or tunable broad-band light source	MLLD
Amplifier at SMF	Broad-band constant or different band gain amplifier	Standard EDFA
Trunk span fiber	SMF	TMF
MUX/DeMUX at RN	AWG	Silicon-on-insulator, PLC, Free-space optics device

	SMF#A	SMF#B
Case A (Both close case)	0 km	0 km
Case B (Single distant case)	0 km	30 km
Case C (Both distant case)	40 km	0 km

	WDM-CDM	MDM-CDM
Light source	Large or tunable broad-band light source	MLLD
Amplifier at SMF	Broad-band constant or different band gain amplifier	Standard EDFA
Trunk span fiber	SMF	TMF
MUX/DeMUX at RN	AWG	Silicon-on-insulator, PLC, Free-space optics device

	SMF#A	SMF#B
Case A (Both close case)	0 km	0 km
Case B (Single distant case)	0 km	30 km
Case C (Both distant case)	40 km	0 km

First demonstration of a scalable MDM/CDM optical access system

Abstract

1. Introduction

2. System architecture

3. Numerical model

4. Experiments

4.1 2 mode x 4 code uplink transmission

4.2 Mode crosstalk tolerance of access span difference

5. Conclusions

Acknowledgments

References and links

Cited By

Figures (7)

Tables (2)

Equations (3)

Optics Express